KR102816308B1 - System, device and method for predicting failure of electric vehicle charger based on artificial intelligence - Google Patents

Abstract

The present invention relates to a system, device, and method for predicting a failure of an electric vehicle charger based on artificial intelligence (AI). The present invention provides a method for predicting a failure of an electric vehicle charger, including: a step in which a failure prediction device collects status information including a failure history, a charging time, a charging voltage, a charging current, a state of charge (SOC) of the electric vehicle, and a charging temperature from an electric vehicle charger that charges an electric vehicle; and a step in which an AI-based failure prediction model uses the status information collected by the failure prediction device as an input variable and a predicted value for the failure time and presence of a failure of the electric vehicle charger as an output variable to predict the status of the electric vehicle charger.

Description

{System, device and method for predicting failure of electric vehicle charger based on artificial intelligence}

The present invention relates to a technology for managing an electric vehicle charger, and more specifically, to a system, device, and method for predicting a failure of an electric vehicle charger based on artificial intelligence (AI).

As electric vehicles become more widespread, the installation of electric vehicle chargers is also expanding. Electric vehicle charging is accomplished through a physical connection between the connector of the electric vehicle charger and the inlet of the electric vehicle.

As with other electronic devices, the lifespan of these electric vehicle charger components increases with increased usage, which increases the likelihood of failure.

If we categorize the types of failures of electric vehicle chargers, they can be categorized into failures (problems that can be temporarily solved by resetting, etc.), breakdowns (failure measures for problems that cannot be solved by resetting), and emergency dispatch (problems that require dispatch within 2 hours). All three types of failures can cause inconvenience to customers. In particular, in the case of emergency dispatch, it is a serious failure that should not occur because it affects the customer's electric vehicle movement, such as the problem of the electric vehicle charger connector not being separated from the electric vehicle inlet.

This problem of connector and inlet mis-separation is caused by the deterioration or aging of electric vehicle chargers due to the rapid increase in the use of electric vehicle chargers. In other words, the deterioration of the electric vehicle charger during charging is causing the heat generated between the connector and inlet to fuse, resulting in the mis-separation problem.

Publication of Patent Publication No. 2023-0123616 (August 24, 2023)

Therefore, in addition to expanding the distribution of electric vehicle chargers, there is a need for technology that can predict when and whether an electric vehicle charger will fail and perform rapid maintenance before the charger fails.

The purpose of the present invention is to provide a system, device, and method for predicting failure of an electric vehicle charger based on artificial intelligence.

Another object of the present invention is to provide an artificial intelligence-based electric vehicle charger failure prediction system, device and method that provide information that enables rapid maintenance to be performed before a failure of the electric vehicle charger occurs.

Meanwhile, the purpose of the present invention is not limited to the above purpose, and other purposes not mentioned can be clearly understood from the description below.

In order to achieve the above object, the present invention provides a method for predicting a failure of an electric vehicle charger, including the steps of: collecting, by a failure prediction device, status information including a failure history, a charging time, a charging voltage, a charging current, a state of charge (SOC) of an electric vehicle, and a charging temperature from an electric vehicle charger that performs charging of an electric vehicle; and predicting the status of the electric vehicle charger using an artificial intelligence-based failure prediction model that uses the status information collected by the failure prediction device as an input variable and a predicted value for the time of failure and presence of failure of the electric vehicle charger as an output variable.

The method for predicting a failure of an electric vehicle charger according to the present invention further includes a step, performed prior to the predicting step, of allowing the failure prediction device to learn the failure prediction model using the status information collected in the past as learning data.

In the above-mentioned predicting step, the fault prediction device can predict the status of the electric vehicle charger using the learned fault prediction model.

In the above collecting step, the fault prediction device stores the status information collected from the electric vehicle charger in a database on an hourly basis.

In the above learning step, the fault prediction device additionally learns the fault prediction model using the status information stored in the database as learning data.

In the collecting step, the charging temperature includes a first internal temperature and a first internal temperature change rate of the electric vehicle charger connected to the inlet of the electric vehicle and charging, and a second internal temperature and a second internal temperature change rate of the electric vehicle charger disconnected from the inlet of the electric vehicle after charging.

In the above-described predicting step, the failure prediction model predicts the degree of deterioration and aging of the electric vehicle charger from a first comparison value compared with the first internal temperature change rate and the first reference temperature change rate, and a second comparison value compared with the second internal temperature change rate and the second reference temperature change rate, and uses the predicted degree of deterioration and aging of the electric vehicle charger as one of the predicted values.

The above first internal temperature change rate is the temperature change rate in the section in which the first internal temperature increases according to the flow of charging time.

The above second internal temperature change rate is the temperature change rate in the section in which the second internal temperature decreases over time after charging is terminated.

The method for predicting a failure of an electric vehicle charger according to the present invention further includes a step of controlling charging of the electric vehicle charger by using the degree of deterioration and aging of the electric vehicle charger predicted by the failure prediction device based on the charging temperature, which is performed after the predicting step.

The above controlling step includes: a step in which, if the first internal temperature of the electric vehicle charger being charged is lower than or equal to a first reference temperature, the fault prediction device maintains the charging current at the first charging current being charged; a step in which, if the first internal temperature of the electric vehicle charger being charged exceeds the first reference temperature, the fault prediction device reduces the charging current to a second charging current lower than the first charging current; and a step in which, if the first internal temperature of the electric vehicle charger being charged exceeds the first reference temperature and is equal to or higher than a set first threshold temperature, the fault prediction device stops charging the electric vehicle.

The above-described predicting step includes a step in which the fault prediction device counts errors for each input variable; and a step in which the fault prediction device predicts the status of the electric vehicle charger at multiple levels based on the number of errors counted.

The above-mentioned predicting step includes a step in which the fault prediction device counts errors for each input variable; and a step in which the fault prediction device predicts the state of the electric vehicle charger based on the number of errors counted by the fault prediction device as a fault level and fault type for each component of the electric vehicle charger.

The method for predicting a failure of an electric vehicle charger according to the present invention further includes a step, performed after the predicting step, in which the failure predicting device outputs information on the maintenance period and failure of the electric vehicle charger using the predicted state of the electric vehicle charger.

The present invention also provides a failure prediction device for an electric vehicle charger, including: a collection unit that collects status information including a failure history, a charging time, a charging voltage, a charging current, a state of charge (SOC) of an electric vehicle, and a charging temperature from an electric vehicle charger that performs charging of an electric vehicle; and a failure prediction unit that predicts the status of the electric vehicle charger using an artificial intelligence-based failure prediction model that uses the collected status information as an input variable and a prediction value regarding the time of failure and presence of failure of the electric vehicle charger as an output variable.

The present invention also provides a failure prediction system for an electric vehicle charger, including: an electric vehicle charger for charging an electric vehicle; and a failure prediction device for predicting a failure of the electric vehicle charger.

The failure prediction system of an electric vehicle charger according to the present invention further includes a database storing the status information collected by the collection unit in units of time; and a management device managing the electric vehicle charger.

The above collection unit is installed in the electric vehicle charger, and the fault prediction unit may be installed in one of the electric vehicle charger and the management device, or may be installed as a separate device.

And, a computer-readable recording medium having recorded thereon a program for performing a method for predicting failure of an electric vehicle charger according to the present invention is provided.

According to the present invention, a failure prediction device can predict a failure of an electric vehicle charger based on artificial intelligence (AI).

This can reduce customer inconvenience caused by failures of electric vehicle chargers by enabling maintenance to be performed on electric vehicle chargers whose failures are predicted based on failure prediction information.

The failure prediction of an electric vehicle charger according to the present invention is performed using an artificial intelligence (AI)-based failure prediction model that uses status information including failure history, charging time, charging voltage, charging current, state of charge (SOC) of the electric vehicle, and charging temperature as input variables, and a predicted value regarding the failure time and presence of failure of the electric vehicle charger as output variables.

The failure prediction of the electric vehicle charger according to the present invention can be provided in multiple failure levels so that the user can easily check the status of the electric vehicle charger. Furthermore, the present invention provides the status of the electric vehicle charger by predicting it in terms of failure levels and failure types for each component of the electric vehicle charger.

The fault prediction of the electric vehicle charger according to the present invention can predict the state of the electric vehicle charger based on the charging temperature among the state information of the electric vehicle charger. That is, the fault prediction device can more accurately predict the degree of deterioration and aging of the electric vehicle charger based on the first internal temperature and the first internal temperature change rate of the electric vehicle charger connected to the inlet of the electric vehicle and charging, and the second internal temperature and the second internal temperature change rate of the electric vehicle charger separated from the inlet of the electric vehicle after charging. As a result, the problem of non-separation due to fusion between the connector and inlet of the electric vehicle charger during charging can be suppressed in advance.

The failure prediction device according to the present invention provides information on the timing and presence of a failure of an electric vehicle charger before a failure occurs in the electric vehicle charger, thereby enabling rapid maintenance to be performed on an electric vehicle charger requiring maintenance.

Meanwhile, the effects of the present invention are not limited to the effects described above, and other effects not mentioned may be directly or implicitly disclosed in the detailed description according to the embodiments of the present invention to be described later.

FIG. 1 is a block diagram showing a failure prediction system for an artificial intelligence-based electric vehicle charger according to an embodiment of the present invention.
Fig. 2 is a block diagram showing the fault prediction device of Fig. 1.
FIG. 3 is a flowchart of a method for predicting failure of an artificial intelligence-based electric vehicle charger according to an embodiment of the present invention.
FIG. 4 is a detailed flowchart for a first example of a step for predicting the status of the electric vehicle charger of FIG. 3.
FIG. 5 is a detailed flowchart for a second example of a step for predicting the status of an electric vehicle charger of FIG. 3.
Figure 6 is a detailed flowchart of the steps for controlling charging of the electric vehicle charger of Figure 3.

It should be noted that in the following description, only the parts necessary for understanding the embodiments of the present invention are described, and the description of other parts will be omitted without departing from the scope of the present invention.

The terms or words used in this specification and claims described below should not be interpreted as limited to their usual or dictionary meanings, but should be interpreted as meanings and concepts that conform to the technical idea of the present invention based on the principle that the inventor can appropriately define the concept of the term in order to describe his own invention in the best way. Therefore, the embodiments described in this specification and the configurations illustrated in the drawings are merely preferred embodiments of the present invention and do not represent all of the technical idea of the present invention, and it should be understood that there may be various equivalents and modified examples that can replace them at the time of filing this application.

Hereinafter, embodiments of the present invention will be described in more detail with reference to the attached drawings.

Fig. 1 is a block diagram showing a failure prediction system for an artificial intelligence-based electric vehicle charger according to an embodiment of the present invention. And Fig. 2 is a block diagram showing a failure prediction device of Fig. 1.

Referring to FIGS. 1 and 2, a failure prediction system (100) of an electric vehicle charger (20) according to the present embodiment is a system that predicts failure of an electric vehicle charger (20) using a failure prediction model based on artificial intelligence (AI).

Here, artificial intelligence (AI) refers to a field that studies artificial intelligence or the methodologies for creating it. Machine learning (machine learning or learning) refers to a field that defines various problems in the field of artificial intelligence and studies the methodologies for solving them. Machine learning is also defined as an algorithm or model that improves the performance of a task through constant experience with the task.

An artificial neural network (ANN) is a model used in machine learning, and can refer to a model with problem-solving capabilities that is composed of artificial neurons (nodes) that form a network by combining synapses. An artificial neural network can be defined by the connection pattern between neurons in different layers, the learning process that updates model parameters, and the activation function that generates output values.

An artificial neural network may include an input layer, an output layer, and optionally one or more hidden layers. Each layer may include one or more neurons, and the artificial neural network may include synapses connecting neurons. In an artificial neural network, each neuron may output a function value of an activation function for input signals, weights, and biases input through synapses.

Model parameters refer to parameters (variables) that are determined through learning, and include the weights of synaptic connections and the biases of neurons. Hyperparameters refer to parameters that must be set before learning in machine learning algorithms, and may include learning rates, number of iterations, mini-batch sizes, and initialization functions.

The purpose of learning an artificial neural network can be seen as determining model parameters that minimize the loss function. The loss function can be used as an indicator to determine the optimal model parameters during the learning process of an artificial neural network.

Machine learning can be classified into supervised learning, unsupervised learning, and reinforcement learning depending on the learning method.

Supervised learning refers to a method of training an artificial neural network when labels for training data are given. The labels can refer to the correct answer (or result) that the artificial neural network should infer when training data is input to the artificial neural network. Unsupervised learning can refer to a method of training an artificial neural network when labels for training data are not given. Reinforcement learning can refer to a learning method that trains an agent defined in a certain environment to select actions or action sequences that maximize cumulative rewards in each state.

Among artificial neural networks, machine learning implemented with a deep neural network (DNN: Deep Neural Network) that includes multiple hidden layers is also called deep learning, and deep learning is a part of machine learning. Hereinafter, machine learning can be used to mean including deep learning, and is described as 'learning'.

The failure prediction system (100) according to this embodiment includes an electric vehicle charger (20) and a failure prediction device (40). The electric vehicle charger (20) performs charging for the electric vehicle (10). And the failure prediction device (40) predicts a failure of the electric vehicle charger (20) using a failure prediction model based on artificial intelligence (AI).

Here, the failure prediction model includes a pattern detection model (PdM) and a statistical/probability model.

Machine learning algorithms applied to pattern detection models include, but are not limited to, Feedforward Neural Network (FNN), Decision Tree (DT), Random Forest (RF), and Support Vector Machine (SVM).

Statistical/probability models include, but are not limited to, Hidden Markov Model (HMM), Bayesian Network (BN), Gaussian Mixture Model (GMM), Extreme Gradient Boosting (XGBoost), Density-based spatial clustering (DBSC), Principal Component Analysis (PCA), K-means, etc.

This failure prediction device (40) includes a collection unit (43) and a failure prediction unit (45). The collection unit (43) collects status information including a failure history, a charging time, a charging voltage, a charging current, a state of charge (SOC) of the electric vehicle (10), and a charging temperature from an electric vehicle charger (20) that performs charging for the electric vehicle (10). Then, the failure prediction unit (45) uses the collected status information as an input variable and a predicted value for the failure time and failure status of the electric vehicle charger (20) as an output variable to predict the status of the electric vehicle charger (20) using an artificial intelligence (AI)-based failure prediction model.

The failure prediction system (100) according to this embodiment may further include a database (50) and a management device (30).

The electric vehicle (10) is equipped with an inlet (11) that can be connected to an electric vehicle charger (20). The electric vehicle (10) receives power from the electric vehicle charger (20) through the inlet (11) and performs charging. The electric vehicle (10) can communicate with the electric vehicle charger (20), management device (30), or failure prediction device (40) via a communication network. Here, the communication network can use short-range communication, power line communication, mobile communication, etc.

The electric vehicle charger (20) is equipped with a connector (21) that can be connected to the inlet (11) of the electric vehicle (10). The electric vehicle charger (20) performs charging for the electric vehicle charger (20) connected through the connector (21).

The electric vehicle charger (20) communicates with the management device (30) and the failure prediction device (40) via a communication network. The electric vehicle charger (20) can communicate with the failure prediction device (40) via the management device (30). The electric vehicle charger (20) can directly communicate with the failure prediction device (40). Here, the communication network can use short-range communication, power line communication, mobile communication, etc.

The electric vehicle charger (20) is equipped with a sensor module that detects and outputs status information including a fault history, charging time, charging voltage, charging current, state of charge (SOC) of the electric vehicle (10), and charging temperature. The sensor module provides the detected status information to a collection unit (43). Alternatively, a sensor module installed in the electric vehicle charger (20) can perform the function of the collection unit (43).

The database (50) stores status information collected from the collection unit (43) on a time basis under the control of the management device (30).

The database (50) stores learning data for learning the failure prediction model. That is, the database (50) preprocesses the status information of the collected electric vehicle charger (20) to generate learning data and stores the generated learning data. The status information of the electric vehicle charger (20) collected here includes status information collected in the past or present from an electric vehicle charger (20) being managed by a management device (30), and may further include status information of other electric vehicle chargers (20) collected through a communication network.

The management device (30) is a device that performs overall management of the electric vehicle charger (20). Here, the management device (30) may include at least one of a local controller and a CSMS (Charging Station Management System).

The management device (30) can store status information collected from the electric vehicle charger (20) in a database (50) on an hourly basis. The management device (30) can transmit status information collected from the electric vehicle charger (20) to a failure prediction device (40).

And the failure prediction device (40) predicts the status of the electric vehicle charger (20) using a failure prediction model based on artificial intelligence (AI). This failure prediction device (40) includes a collection unit (43) and a failure prediction unit (45) as described above, and may further include a communication unit (41) and a storage unit (47).

The communication unit (41) communicates with the electric vehicle charger (20) and the management device (30). The communication unit (41) can receive status information of the electric vehicle charger (20) from the electric vehicle charger (20) and transmit it to the failure prediction unit (45). Alternatively, the communication unit (41) can receive status information of the electric vehicle charger (20) from the management device (30).

The collection unit (43) collects status information of the electric vehicle charger (20). The collection unit (43) can store the collected status information in a database (50) by time unit.

This collection unit (43) can collect status information from an electric vehicle charger (20) through a communication unit (41) and transmit the collected status information to a failure prediction unit (45).

This collection unit (43) may be installed in a separate failure prediction device (40) as shown in Fig. 1, but is not limited thereto. For example, the collection unit (43) may be installed in an electric vehicle charger (20) or a management device (30).

Alternatively, the collection unit (43) can transmit the collected status information to the management device (30) via the communication unit (41). The management device (30) can store the received status information in the database (50). The management device (30) can transmit the received status information to the failure prediction unit (45).

The storage unit (47) stores a program required for controlling the operation of the failure prediction device (40) and information generated during the execution of the program. The storage unit (47) stores a failure prediction model that performs failure prediction of the electric vehicle charger (20). The storage unit (47) can temporarily store status information collected by the collection unit (43). The storage unit (47) stores the status of the electric vehicle charger (20) predicted by the failure prediction model, that is, predicted values for the failure time and failure status of the electric vehicle charger (20). The storage (47) may include at least one storage medium among a flash memory type, a hard disk type, a multimedia card micro type, a card type memory (e.g., an SD or XD memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, and an optical disk.

And the failure prediction unit (45) includes at least one processor that performs failure prediction of the electric vehicle charger (20).

The fault prediction unit (45) performs learning on a fault prediction model that performs fault prediction of an electric vehicle charger (20). That is, the fault prediction unit (45) trains the fault prediction model using status information of the electric vehicle charger (20) collected in the past as learning data. The fault prediction unit (45) additionally trains the fault prediction model using status information of the electric vehicle charger (20) currently collected as learning data. Here, the learning data is provided by a database (50).

The failure prediction unit (45) uses the status information of the electric vehicle charger (20) collected by the collection unit (43) as an input variable and predicts the status of the electric vehicle charger (20) using a learned failure prediction model that uses the predicted values for the failure time and failure status of the electric vehicle charger (20) as output variables.

This fault prediction unit (45) counts errors for each input variable and can predict the status of the electric vehicle charger (20) at multiple levels based on the counted number of errors. For example, the fault prediction unit (45) can predict at 1st to 4th levels as shown in Table 1 below.

Level explanation 1st place 1. Safety issues
- Signs of electrical faults, sparks or overheating
- Damaged or exposed wiring and connectors
- Malfunctioning emergency stop mechanism
2. Major failure
- If the charger does not turn on or cannot charge the vehicle
- Sudden decrease or complete cessation of charging performance
3. Security Vulnerabilities
- Outdated firmware or software with known security vulnerabilities
- Unauthorized access or tampering with the charger 2nd place 1. Parts wear
- Signs of wear and tear on cables and connectors that could lead to damage
- Deterioration of insulation or minor corrosion of electrical components
2. Intermittent problems
- Charging interruption or irregular charging behavior
- Displays error messages or warnings that indicate potential problems but do not stop charging
3. Environmental protection:
- Damaged seals or enclosures that allow moisture or dust to enter. 3rd place 1. Preventive maintenance
- Regular inspection and cleaning
- Testing and recalibrating the system
- Firmware and software updates
2. Minor repairs
- Replace worn parts that are still functional
- Minor corrosion or rust treatment.
3. Performance Optimization
- Ensures the charger operates at optimal efficiency
- Identify minor abnormalities through performance data analysis 4th place 1. Non-core appearance issues
- Minor scratches, dents or cosmetic damage that do not affect functionality
- Cleaning and minor adjustments
2. Documentation and record keeping
- Update maintenance logs and records
- Review historical data for long-term maintenance planning
3. Education and Audit
- Conduct maintenance personnel training sessions
- Conduct safety audits and compliance checks

Here, the first priority is a failure that requires immediate action, the second priority is a failure that requires action within a week, the third priority is a failure that requires action on a monthly basis, and the fourth priority is a failure that requires action on a quarterly/annual basis.

Alternatively, the fault prediction unit (45) can count errors by input variable and predict the state of the electric vehicle charger (20) based on the counted number of errors as a fault level and fault type for each component of the electric vehicle charger (20). Here, the electric vehicle charger (20) can include an input unit, an output unit, a communication unit, a control unit, and other function performing units. For example, the fault prediction unit (45) can predict the state of the electric vehicle charger (20) as shown in Table 2 below.

Table 2 classifies failure levels into levels 1 to 3, but is not limited to these.

This fault prediction unit (45) can predict the status of the electric vehicle charger (20) based on the charging temperature among the status information of the electric vehicle charger (20) collected by the collection unit (43). The charging temperature includes the first internal temperature and the first internal temperature change rate of the electric vehicle charger (20) connected to the inlet (11) of the electric vehicle (10) and being charged, and the second internal temperature and the second internal temperature change rate of the electric vehicle charger (20) disconnected from the inlet (11) of the electric vehicle (10) after charging.

The fault prediction unit (45) predicts the degree of deterioration and aging of the electric vehicle charger (20) from a first comparison value obtained by comparing the first internal temperature change rate with the first reference temperature change rate and a second comparison value obtained by comparing the second internal temperature change rate with the second reference temperature change rate using a fault prediction model. The fault prediction unit (45) uses the predicted degree of deterioration and aging of the electric vehicle charger (20) as one of the predicted values.

The first internal temperature change rate is the temperature change rate in the section where the first internal temperature increases over the passage of charging time. The second internal temperature change rate is the temperature change rate in the section where the second internal temperature decreases over the passage of time after charging is completed.

The reason why the failure prediction unit (45) according to the present embodiment uses the first and second internal temperature change rates as input variables in addition to the first and second internal temperatures is as follows.

The first internal temperature provides information about the current status of the electric vehicle charger (20) while charging, but does not provide information that can predict the degree of deterioration and aging of the electric vehicle charger (20).

However, since the degree of deterioration and aging of the electric vehicle charger (20) is proportional to the service life of the electric vehicle charger (20), the deterioration and aging of the electric vehicle charger (20) progresses as the charging time of the electric vehicle charger (20) increases. The degree of deterioration and aging of the electric vehicle charger (20) is correlated with the first internal temperature and the first internal temperature change rate, and the second internal temperature and the second internal temperature change rate.

Therefore, the failure prediction unit (45) according to the present embodiment predicted the degree of deterioration and aging of the electric vehicle charger (20) from the first internal temperature and the first internal temperature change rate, and the second internal temperature and the second internal temperature change rate. And since the degree of deterioration and aging of the electric vehicle charger (20) is more related to the first internal temperature and the first internal temperature change rate than to the second internal temperature and the second internal temperature change rate, a relatively higher weight can be set to the first input variable for the first internal temperature and the first internal temperature change rate than to the second input variable for the second internal temperature and the second internal temperature change rate.

The failure prediction unit (45) can use the predicted status of the electric vehicle charger (20) to provide information on the maintenance period and failure of the electric vehicle charger (20) to at least one of the electric vehicle charger (20) and the management device (30).

The management device (30) manages information on the maintenance period and failure of the received electric vehicle charger (20), and can provide work instructions to workers for electric vehicle chargers (20) requiring maintenance and failure based on the received information.

The electric vehicle charger (20) can notify the surrounding area of the electric vehicle charger (20) of a warning, alarm, etc. through a display, speaker, warning light, etc. equipped on the electric vehicle charger (20) when the maintenance period has arrived or a failure has occurred, based on the information on the maintenance period and failure of the electric vehicle charger (20) received.

If the predicted state of the electric vehicle charger (20) is a risk of failure or fire, one of the failure prediction unit (45), the electric vehicle charger (20), and the management device (30) can notify the state of the electric vehicle charger (20) to the customer terminal of the electric vehicle (10) being charged in the electric vehicle charger (20). Here, the customer terminal includes at least one of an infotainment system equipped in the electric vehicle (10) and a communication terminal carried by the customer.

The failure prediction unit (45) can control the charging of the electric vehicle charger (20) as follows by using the degree of deterioration and aging of the electric vehicle charger (20) predicted based on the charging temperature.

That is, when the first internal temperature of the electric vehicle charger (20) being charged is lower than the first reference temperature, the fault prediction unit (45) can provide a first control signal to the electric vehicle charger (20) to maintain the charging current at the first charging current currently being charged.

If the first internal temperature of the electric vehicle charger (20) during charging exceeds the first reference temperature, the fault prediction unit (45) can provide a second control signal to the electric vehicle charger (20) to reduce the charging current to a second charging current lower than the first charging current.

And when the first internal temperature of the electric vehicle charger (20) being charged exceeds the first reference temperature and is equal to or higher than the set first threshold temperature, the fault prediction unit (45) can provide a third control signal to the electric vehicle charger (20) to stop charging the electric vehicle (10).

The failure prediction device (40) according to this embodiment may be implemented as a separate device from the electric vehicle charger (20) and the management device (30), as shown in FIG. 1, but is not limited thereto. For example, the failure prediction device (40) may be embedded in one of the electric vehicle charger (20) and the management device (30). Alternatively, the collection unit (43) of the failure prediction device (40) may be installed in the electric vehicle charger (20), and the failure prediction unit (45) of the failure prediction device (40) may be installed in one of the management devices (30).

The method for predicting a failure of an electric vehicle charger (20) using a failure prediction system (100) according to this embodiment will be described with reference to FIGS. 1 to 3. Here, FIG. 3 is a flow chart of a method for predicting a failure of an electric vehicle charger (20) based on artificial intelligence according to an embodiment of the present invention.

First, in step S10, the fault prediction device (40) trains a fault prediction model using status information including fault history, charging time, charging voltage, charging current, state of charge (SOC) of the electric vehicle (10) and charging temperature collected from an electric vehicle charger (20) that performed charging in the past as learning data.

Next, in step S20, the fault prediction device (40) collects status information including fault history, charging time, charging voltage, charging current, state of charge (SOC) of the electric vehicle (10), and charging temperature from the electric vehicle charger (20) that performs charging for the electric vehicle (10).

Next, in step S30, the failure prediction device (40) predicts the status of the electric vehicle charger (20) using an artificial intelligence-based failure prediction model that uses the collected status information as input variables and the predicted values for the failure time and failure status of the electric vehicle charger (20) as output variables.

Next, in step S40, the failure prediction device (40) can output information on the maintenance period and failure of the electric vehicle charger (20) using the predicted state of the electric vehicle charger (20). For example, the failure prediction device (40) can output information on the maintenance period and failure of the electric vehicle charger (20) to at least one of the electric vehicle charger (20), the management device (30), and the customer terminal of the electric vehicle (10) being charged by the electric vehicle charger (20).

And at step S50, the failure prediction device (40) can control charging of the electric vehicle charger (20) by using the degree of deterioration and aging of the electric vehicle charger (20) predicted based on the charging temperature.

The steps for predicting the state of the electric vehicle charger (20) according to the S30 step are described below with reference to FIGS. 1 to 5. Here, FIG. 4 is a detailed flowchart for a first example of the steps for predicting the state of the electric vehicle charger (20) of FIG. 3. And FIG. 5 is a detailed flowchart for a second example of the steps for predicting the state of the electric vehicle charger (20) of FIG. 3.

First, referring to FIGS. 1 to 4, a first example of a step for predicting the state of an electric vehicle charger (20) according to step S30 will be described as follows.

First, in step S31, the fault prediction device (40) counts errors for each input variable.

And in step S33, the fault prediction device (40) predicts the status of the electric vehicle charger (20) at multiple levels based on the counted number of errors.

First, a second example of a step of predicting the state of an electric vehicle charger (20) according to step S30 will be described with reference to FIGS. 1 to 3 and FIG. 5.

First, in step S31, the fault prediction device (40) counts errors for each input variable.

And at step S35, the failure prediction device (40) can predict the status of the electric vehicle charger (20) based on the counted number of errors, the failure level and failure type of each component of the electric vehicle charger (20).

The steps for controlling charging of an electric vehicle charger (20) according to the S50 step are described below with reference to FIGS. 1 to 3 and FIG. 6. Here, FIG. 6 is a detailed flowchart for the steps for controlling charging of an electric vehicle charger (20) of FIG. 3.

First, in step S51, the fault prediction device (40) maintains the charging current at the first charging current that is currently being charged if the first internal temperature of the electric vehicle charger (20) that is being charged is lower than or equal to the first reference temperature. That is, the fault prediction device (40) can provide the electric vehicle charger (20) with a first control signal that maintains the charging current at the first charging current that is currently being charged.

Next, in step S53, the fault prediction device (40) reduces the charging current to a second charging current lower than the first charging current if the first internal temperature of the electric vehicle charger (20) being charged exceeds the first reference temperature. That is, the fault prediction device (40) can provide a second control signal to the electric vehicle charger (20) that reduces the charging current to a second charging current lower than the first charging current.

And in step S55, the fault prediction device (40) stops charging the electric vehicle (10) if the first internal temperature of the electric vehicle charger (20) being charged exceeds the first reference temperature and is equal to or higher than the set first threshold temperature. That is, the fault prediction device (40) can provide a third control signal to the electric vehicle charger (20) to stop charging the electric vehicle (10).

Although the present embodiment discloses an example of performing step S50 after step S40, it is not limited to this. For example, the failure prediction device (40) may perform step S40 after step S50, or may perform steps S40 and S50 simultaneously.

Meanwhile, the method for predicting a failure of an electric vehicle charger (20) according to the above-described embodiment of the present invention may be implemented in the form of a program that can be read by various computer means and recorded on a computer-readable recording medium. Here, the recording medium may include program commands, data files, data structures, etc., alone or in combination. The program commands recorded on the recording medium may be those specially designed and configured for the present invention or may be those known to and usable by those skilled in the art of computer software. For example, the recording medium includes magnetic media such as hard disks, floppy disks, and magnetic tapes, optical media such as CD-ROMs and DVDs, magneto-optical media such as floptical disks, and hardware devices specially configured to store and execute program commands such as ROMs, RAMs, and flash memories. Examples of program commands may include not only machine language wires generated by a compiler, but also high-level language wires that can be executed by a computer using an interpreter, etc. Such hardware devices may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

Meanwhile, the embodiments disclosed in this specification and drawings are merely specific examples presented to aid understanding and are not intended to limit the scope of the present invention. It will be apparent to those skilled in the art to which the present invention pertains that other modified examples based on the technical idea of the present invention can be implemented in addition to the embodiments disclosed herein.

10: Electric car 11: Inlet
20 : Electric vehicle charger 21 : Connector
30: Management device 40: Failure prediction device
41: Communication Department 43: Collection Department
45: Fault Prediction Unit 47: Storage Unit
50: Database 100: Failure Prediction System

Claims (12)

A step of collecting status information including a fault history, a charging time, a charging voltage, a charging current, a state of charge (SOC) of the electric vehicle, and a charging temperature from an electric vehicle charger that performs charging for the electric vehicle by a fault prediction device; and
A step of predicting the status of the electric vehicle charger using an artificial intelligence-based failure prediction model that uses the status information collected by the failure prediction device as an input variable and a predicted value regarding the time of failure and whether the electric vehicle charger has failed as an output variable; including;
In the above collecting step,
The above charging temperature includes a first internal temperature and a first internal temperature change rate of the electric vehicle charger connected to the inlet of the electric vehicle and charging, and a second internal temperature and a second internal temperature change rate of the electric vehicle charger disconnected from the inlet of the electric vehicle after charging.
In the above prediction step,
A method for predicting a failure of an electric vehicle charger, characterized in that the above failure prediction model predicts the degree of deterioration and aging of the electric vehicle charger from a first comparison value compared with the first internal temperature change rate and the first reference temperature change rate, and a second comparison value compared with the second internal temperature change rate and the second reference temperature change rate, and uses the predicted degree of deterioration and aging of the electric vehicle charger as one of the predicted values. In the first paragraph, performed before the predicting step,
The above fault prediction device further includes a step of training the fault prediction model using the status information collected in the past as training data;
In the above prediction step,
A method for predicting a failure of an electric vehicle charger, characterized in that the failure prediction device predicts the state of the electric vehicle charger using the learned failure prediction model. In the second paragraph, in the collecting step,
The above fault prediction device stores the status information collected from the electric vehicle charger in a database on an hourly basis,
In the above learning step,
A method for predicting a failure of an electric vehicle charger, characterized in that the failure prediction device additionally learns the failure prediction model using the status information stored in the database as learning data. In the first paragraph,
In the above prediction step,
A method for predicting a failure of an electric vehicle charger, characterized in that a relatively higher weight is set to a first input variable for the first internal temperature and the first internal temperature change rate than to a second input variable for the second internal temperature and the second internal temperature change rate. In the first paragraph,
The above first internal temperature change rate is the temperature change rate in the section where the first internal temperature increases according to the flow of charging time,
A method for predicting a failure of an electric vehicle charger, characterized in that the second internal temperature change rate is a temperature change rate in a section in which the second internal temperature decreases over time after charging is terminated. In the first paragraph, performed after the predicting step,
The above fault prediction device further includes a step of controlling charging of the electric vehicle charger by using the degree of deterioration and aging of the electric vehicle charger predicted based on the charging temperature;
The above controlling step is,
A step for the fault prediction device to maintain the charging current at the first charging current when the first internal temperature of the electric vehicle charger being charged is lower than or equal to the first reference temperature;
When the first internal temperature of the electric vehicle charger during charging exceeds the first reference temperature, the fault prediction device reduces the charging current to a second charging current lower than the first charging current; and
A step for the fault prediction device to stop charging the electric vehicle when the first internal temperature of the electric vehicle charger during charging exceeds the first reference temperature and is equal to or higher than the set first threshold temperature;
A method for predicting a failure of an electric vehicle charger, characterized by including a . In the first paragraph, the predicting step,
a step in which the above fault prediction device counts errors for each input variable; and
A step of predicting the status of the electric vehicle charger into multiple levels based on the number of errors counted by the above fault prediction device;
which is performed after the above predicting step,
The above failure prediction device outputs information on the maintenance period and failure of the electric vehicle charger using the predicted status of the electric vehicle charger;
A method for predicting a failure of an electric vehicle charger, characterized in that it further includes: In the first paragraph, the predicting step,
a step in which the above fault prediction device counts errors for each input variable; and
A step of predicting the status of the electric vehicle charger based on the number of errors counted by the above-mentioned fault prediction device as a fault level and fault type for each component of the electric vehicle charger;
which is performed after the above predicting step,
The above failure prediction device outputs information on the maintenance period and failure of the electric vehicle charger using the predicted status of the electric vehicle charger;
A method for predicting a failure of an electric vehicle charger, characterized in that it further includes: A collection unit that collects status information including fault history, charging time, charging voltage, charging current, state of charge (SOC) of the electric vehicle, and charging temperature from an electric vehicle charger that performs charging of the electric vehicle; and
A failure prediction unit that predicts the status of the electric vehicle charger using an artificial intelligence-based failure prediction model that uses the collected status information as input variables and a prediction value for the time and presence of failure of the electric vehicle charger as output variables;
The above charging temperature includes a first internal temperature and a first internal temperature change rate of the electric vehicle charger connected to the inlet of the electric vehicle and charging, and a second internal temperature and a second internal temperature change rate of the electric vehicle charger disconnected from the inlet of the electric vehicle after charging.
The above failure prediction model is a failure prediction device for an electric vehicle charger, characterized in that it predicts the degree of deterioration and aging of the electric vehicle charger from a first comparison value compared with the first internal temperature change rate and the first reference temperature change rate, and a second comparison value compared with the second internal temperature change rate and the second reference temperature change rate, and uses the predicted degree of deterioration and aging of the electric vehicle charger as one of the predicted values. An electric vehicle charger for charging electric vehicles; and
A fault prediction device for predicting a fault of the electric vehicle charger;
The above fault prediction device,
A collection unit that collects status information including fault history, charging time, charging voltage, charging current, state of charge (SOC) of the electric vehicle, and charging temperature from an electric vehicle charger that performs charging of the electric vehicle; and
A failure prediction unit that predicts the status of the electric vehicle charger using an artificial intelligence-based failure prediction model that uses the collected status information as input variables and a prediction value for the time and presence of failure of the electric vehicle charger as output variables;
The above charging temperature includes a first internal temperature and a first internal temperature change rate of the electric vehicle charger connected to the inlet of the electric vehicle and charging, and a second internal temperature and a second internal temperature change rate of the electric vehicle charger disconnected from the inlet of the electric vehicle after charging.
The above failure prediction model is a failure prediction system for an electric vehicle charger, characterized in that it predicts the degree of deterioration and aging of the electric vehicle charger from a first comparison value compared with the first internal temperature change rate and the first reference temperature change rate, and a second comparison value compared with the second internal temperature change rate and the second reference temperature change rate, and uses the predicted degree of deterioration and aging of the electric vehicle charger as one of the predicted values. In Article 10,
A database that stores the status information collected from the above collection unit in units of time; and
Further comprising a management device for managing the electric vehicle charger;
The above collection unit is installed in the electric vehicle charger,
A failure prediction system for an electric vehicle charger, characterized in that the above failure prediction unit is installed in one of the electric vehicle charger and the management device, or is installed as a separate device. A computer-readable recording medium having recorded thereon a program for performing a method for predicting a failure of an electric vehicle charger according to any one of claims 1 to 8.

Images