CN116407115A - Abnormality detection method, abnormality detection system and lower limb rehabilitation training device - Google Patents

Abstract

The present application relates to an abnormality detection method, system and lower limb rehabilitation training equipment. The abnormality detection method is applied to rehabilitation training equipment that provides lower limb rehabilitation training for training objects, including: obtaining the abnormality detection level; obtaining the training data of the training object; determining the corresponding abnormality detection strategy according to the abnormality detection level; according to the determined abnormality detection strategy And the training data to determine whether the training object is abnormal. Adopting the abnormality detection method can improve the accuracy of abnormality detection in the process of rehabilitation training.

Description

Abnormality detection method, system and lower limb rehabilitation training equipment

technical field

The present application relates to the technical field of rehabilitation training, in particular to an abnormality detection method, system and lower limb rehabilitation training equipment.

Background technique

With the development of rehabilitation training technology, various rehabilitation training equipments have appeared. In the process of sports rehabilitation training based on rehabilitation training equipment, abnormal spasm may be induced in the trained part of the training object, and the untimely detection or inappropriate treatment of the abnormal spasm will cause secondary injury to the training object.

In the process of sports rehabilitation training, traditional rehabilitation training equipment only uses a simple single threshold to judge whether the training object has abnormalities such as spasticity. However, different training objects or the same training object in different training stages may have abnormalities. Therefore, traditional sports rehabilitation training equipment is not accurate in detecting abnormalities of training objects, and misjudgments or misjudgments often occur.

Contents of the invention

Based on this, it is necessary to provide an abnormality detection method, system and lower limb rehabilitation training equipment capable of improving the accuracy of abnormality detection in the rehabilitation training process in view of the above technical problems.

An abnormality detection method, which is applied to rehabilitation training equipment that provides lower limb rehabilitation training for training objects, comprising: obtaining the abnormality detection level; obtaining training data of the training object; determining the corresponding abnormality detection strategy according to the abnormality detection level; according to the determined abnormality detection Strategy and training data to determine whether the training object is abnormal.

In one embodiment, the anomaly detection strategy includes an anomaly detection model and/or an anomaly judgment condition.

In one embodiment, the anomaly detection model includes evaluating features and/or evaluating relationships.

In one embodiment, the method further includes: acquiring data selection rules; selecting data from training data according to the data selection rules as abnormality detection data; judging whether the training object is abnormal according to the determined abnormality detection strategy and abnormality detection data.

In one embodiment, the data selection rule is to select data according to at least one of anomaly detection level, training parameters and object parameters.

In an embodiment, the method further includes: judging that an abnormality occurs to the training object, and controlling the rehabilitation training equipment to start protection measures.

In one embodiment, the protection measures include at least one of emergency stop, deceleration, switching training mode, warning and reverse.

In one embodiment, there is a corresponding relationship between the abnormality detection level and at least one training mode of the rehabilitation training equipment.

In one embodiment, the method further includes: dynamically adjusting the anomaly detection level during the training process in response to the level adjustment instruction.

An abnormality detection system applied to rehabilitation training equipment that provides lower limb rehabilitation training for training objects, including:

The level acquisition component is used to obtain the abnormality detection level;

The data acquisition component is used to acquire the training data of the training object;

A processor, the processor is connected with the level acquisition component and the data acquisition component, and the processor is used to execute any one of the abnormality detection methods mentioned above.

A lower limb rehabilitation training device, comprising: a training component and the above-mentioned abnormality detection system.

The above abnormality detection method, system and lower limb rehabilitation training equipment determine the corresponding abnormality detection strategy according to the obtained abnormality detection level during the rehabilitation training process of the training object, and combine the training data of the training object according to the determined abnormality detection strategy. Check whether the object is abnormal. Since the anomaly detection strategy is determined based on the anomaly detection level of the training object, for the training objects of different anomaly detection levels, the anomaly detection strategy that matches their anomaly detection level can be determined respectively, and even for the same training object Objects, in their different training stages, as long as the anomaly detection levels are set differently, different detection results can be given, so that blind single detection can be avoided, and the accuracy of anomaly detection can be improved.

Description of drawings

Fig. 1 is a schematic structural diagram of an abnormality detection system in an embodiment;

Fig. 2 is a schematic flow chart of an anomaly detection method in an embodiment;

Fig. 3 is a schematic structural diagram of an abnormality detection device in an embodiment;

Fig. 4 is a structural schematic diagram of lower limb rehabilitation training equipment in an embodiment;

Fig. 5 is a schematic structural diagram of the training components of the lower limb rehabilitation training device in one embodiment.

Detailed ways

In order to make the purpose, technical solution and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present application, not to limit the present application.

Referring to FIG. 1 , FIG. 1 is a schematic structural diagram of an anomaly detection system in an embodiment. Exemplarily, the anomaly detection system 100 provided in the present application may include: a grade acquisition component 102 , a data acquisition component 104 and a processor 106 . Each component in the anomaly detection system can be fully or partially realized by software, hardware and a combination thereof. The above-mentioned components can be embedded in or independent of the processor in the form of hardware, and can also be stored in the memory of the computer device in the form of software, so that the processor can invoke and execute the corresponding operations of the above-mentioned components.

Wherein, the level acquiring component 102 can be used to acquire the abnormality detection level. The level acquiring component 102 may be an external input device, a touch screen, or a data interface, etc., to acquire the abnormality detection level input by the user or preset by the system.

The data acquisition component 104 can be used to acquire training data of the training object. Wherein, the data acquisition component 104 may include at least one sensor. Of course, the data acquisition component 104 may also include a receiving module, a data interface, etc., to acquire training data of the training object from sensors or other detection components.

Exemplarily, the sensor can be set on the training component of the rehabilitation training equipment, or can be directly worn on the lower limbs of the training object. By arranging the sensor, it is possible to dynamically capture training data representing the motion state of the training object during the training process of the training object.

A processor 106, which can be connected with the level acquisition component 102 and the data acquisition component 104, and can communicate. The processor 106 may be a microprocessor, that is, a microcomputer system, or may be a CPU (Central Processing Unit, central processing unit) or the like. The processor 106 may be a processor in any computer device, or a processor in a lower limb rehabilitation training device. There is no limitation to the manner in which the processor 106 communicates with the grade acquisition component 102 and the data acquisition component 104 , and it may be through electrical connection for signal transmission, or through network, interface, etc. for network communication. The connection mode between the processor 106 and the grade acquisition component 102 and the data acquisition component 104 is also not limited, and may be electrically connected, or connected through a network, an interface, and the like.

In an embodiment, reference may be made to FIG. 2 , which shows a schematic flowchart of an anomaly detection method in an embodiment. In this embodiment, the application of the anomaly detection method to the processor 106 of the anomaly detection system 100 is taken as an example for illustration. Specifically, the following steps may be included:

Step S202: Obtain an abnormality detection level.

Wherein, the abnormality detection level refers to the level used to evaluate the possibility of abnormal occurrence of the training object during exercise. For example, the abnormality detection level may be a level used to evaluate the possibility of the occurrence of convulsions during exercise of the training subject. The exercise ability and the abnormality detection level can be set to be inversely proportional, that is, the weaker the exercise ability of the training object, the higher the abnormality detection level. Of course, according to the specific system design, the exercise ability and the abnormality detection level can be set to be proportional, etc.

Exemplarily, the abnormality detection level can be customized according to the exercise ability of the training object, and can also be set with reference to the spasticity level of the training object evaluated by the spasticity scale. Of course, you can also combine the two.

In this step, for example, the user can directly input the anomaly detection that characterizes the possibility of abnormal occurrence of the current training object through the level acquisition component 102, such as an external input device such as a mouse, a keyboard, a button, an adjustment knob, or a touch screen. level, the processor 106 can communicate with the level acquisition component 102 to obtain the abnormality detection level input by the user through the level acquisition component 102; or, the processor 106 can also control the display or touch screen to display selectable abnormalities on the display interface detection level, and in response to the user's selection operation on the abnormality detection level, acquire the abnormality detection level selected by the user.

Step S204: Obtain training data of the training object.

Wherein, the training data is the data representing the motion state of the training object generated during the rehabilitation training process of the training object. Exemplarily, the training data may include relevant data reflecting the motion and/or force of the training object during the training process, for example, may include but not limited to joint angle, joint angular velocity, joint angular acceleration, joint or gait phase, joint interaction One or more of torque and other data.

In this step, for example, the processor 106 may communicate with the data acquisition component 104 to acquire the training data of the training object collected by the data acquisition component 102 in real time or periodically during the training process of the training object.

Step S206: Determine the corresponding anomaly detection strategy according to the anomaly detection level.

Wherein, the anomaly detection strategy refers to a detection strategy for detecting whether an abnormality occurs in a training object. An anomaly detection strategy may include an anomaly detection model and/or an anomaly judgment condition. An anomaly detection strategy can have a corresponding relationship with an anomaly detection level. An anomaly detection level can correspond to at least one anomaly detection strategy. In the case of an anomaly detection level corresponding to multiple anomaly detection strategies, a selection strategy can be further provided for the user An anomaly detection strategy can be selected from a plurality of anomaly detection strategies. Different anomaly detection levels can also be configured with different anomaly detection strategies. The correspondence between anomaly detection levels and anomaly detection strategies can be customized and adjusted according to business requirements and specific conditions of training objects.

Exemplarily, the corresponding relationship between the abnormality detection strategy and the abnormality detection level may be preset by the operator and stored in the memory; it may also be that when the rehabilitation training equipment is started, the processor 106 analyzes the historical training of the training object Data, dynamically mapped and established corresponding relationships, or manually selected or configured corresponding relationships by operators.

In this step, the processor 106 may determine the abnormality detection strategy corresponding to the abnormality detection level according to the acquired abnormality detection level and based on the pre-configured correspondence between the abnormality detection level and the abnormality detection strategy.

Step S208: According to the determined abnormality detection strategy and training data, determine whether the training object is abnormal.

In this step, the processor 106 may determine whether the training object is abnormal according to the acquired training data of the training object and the determined abnormality detection strategy corresponding to the abnormality detection level. Optionally, the processor 106 may use all or part of the training data as data for anomaly detection.

In the above-mentioned anomaly detection method, during the rehabilitation training process of the training object, the corresponding anomaly detection strategy is determined according to the acquired anomaly detection level, and whether the abnormality occurs in the training object is judged according to the determined anomaly detection strategy combined with the training data of the training object . Since the anomaly detection strategy is determined based on the anomaly detection level of the training object, for the training objects of different anomaly detection levels, the anomaly detection strategy that matches their anomaly detection level can be determined respectively, and, even for the same As long as the anomaly detection level is set to be different, different detection results can be given for the training object in different training stages, so that blind single detection can be avoided and the accuracy of anomaly detection can be improved.

In one embodiment, an anomaly detection strategy may include an anomaly detection model.

Wherein, the abnormality detection model is a calculation model for evaluating whether the training object is abnormal during the training process, for example, it may be a model for evaluating the possibility of the training object having convulsions during the training process.

Exemplarily, the anomaly detection model can use any one or more types of data in the training data set, reference training data set, and training parameters as the characteristic variables associated with the model for calculation to obtain the anomaly detection result. Moreover, the functional correlation between the feature variables of the anomaly detection model can also be customized and adjusted as needed.

In one embodiment, an anomaly detection model may include evaluation features. Among them, the evaluation features refer to the characteristic variables of the anomaly detection model used for anomaly evaluation. In this embodiment, the processor 106 may determine the evaluation features used for anomaly detection according to the obtained anomaly detection level.

Exemplarily, any one or more types of data in the training data set, reference data set and training parameters can be used as the evaluation feature. The training data set may include training data of the training object, for example, joint angles, joint angular velocities, joint angular accelerations, joint or gait phases, or joint interaction torques. The reference training data set may include reference data reflecting the dynamic relationship of the data calculated according to the physical parameters of the training object or the historical training data of the training object, such as reference joint torque and the like. The training parameters may include training related parameters such as joint angle range, walking stride, walking frequency, walking speed, and walking duration. According to specific needs, the evaluation feature can be determined from any type of data or a combination of multiple types of data in the above training data set, reference data set and training parameters.

Exemplarily, the abnormality detection level can also be used as one of the evaluation features of the abnormality detection model. Of course, the combination of any one or more types of data in the training data set, reference data set and training parameters and the abnormality detection level can also be used as the evaluation feature of the abnormality detection model as required.

Since the types and/or quantities of evaluation features determined by different anomaly detection levels can be different, even for the same training object, the training data set includes the same data, and the selected evaluation features are different, giving Different anomaly detection results. In addition, the anomaly detection level itself can also be used as an evaluation feature, so different anomaly detection levels can give different anomaly detection results. Further, the object parameters and/or the dynamic relationship data established according to the historical training data of the training object are set as feature variables, therefore, for different training objects, different anomaly detection can be given according to the object parameters and object historical data result.

In one embodiment, an anomaly detection model may include evaluating relationships. Wherein, the evaluation relationship refers to a functional correlation between evaluation features. In this embodiment, the processor 106 may determine an evaluation relationship among evaluation features used for abnormal detection according to the obtained abnormal detection level.

Exemplarily, the evaluation relationship among the evaluation features of the abnormality detection model may be determined based on a dynamic model and/or a kinematics model according to business requirements. Different anomaly detection levels may correspond to different evaluation relationships. Therefore, even if the same evaluation features are used, different anomaly detection results will be given due to different evaluation relationships among the evaluation features, that is, different functional correlations. Further, different training parameters may also correspond to different evaluation relationships.

In one embodiment, an anomaly detection model may include evaluating features and evaluating relationships.

In this embodiment, the processor 106 may determine any combination of one or more types of data in the above-mentioned training data set, reference data set, and training parameters as the evaluation feature according to the acquired abnormality detection level, or, the above-mentioned The combination of any one or more types of data in the training data set, reference data set and training parameters and the level of abnormality detection is determined as the evaluation feature, or only the level of abnormality detection is determined as the evaluation feature, etc., and any A set of functional associations serves as evaluation relations between evaluation features. Since different anomaly detection levels correspond to different evaluation features and evaluation relationships, different anomaly detection results can be given for different anomaly detection levels.

In one embodiment, the anomaly detection strategy may include anomaly judgment conditions. Wherein, the abnormality judging condition refers to a condition for judging whether the training object is abnormal.

Exemplarily, the abnormal judgment conditions may include but not limited to: whether the motion state data of one or more joints of the training object exceed respective thresholds; whether the motion state data of one or more joints of the training object exceed their respective ranges Value limit; at a certain point in time or in a period of time, whether the proportion of the exercise state data of the training object exceeds the respective threshold reaches a certain preset ratio; at a certain point in time or in a period of time, whether the rate of change of the exercise state data of the training object exceeds Preset standard; whether the data change range between the exercise state data sampled at the current time point or time period and the exercise state data sampled at the preset time point or time period exceeds the preset standard, etc.

Exemplarily, it is possible to set corresponding abnormality judgment conditions for different abnormality detection levels, that is, to increase or decrease the abnormality judgment judgment conditions corresponding to the high and low changes of the abnormality detection level, so that the abnormality detection The system is more sensitive to the detection of training objects with weak athletic ability, and can correspondingly relax the judgment standard for training objects with strong athletic ability, thereby reducing the occurrence of false detection.

Further, the abnormal judgment condition may also be related to the training parameters. For example, the training parameters such as joint angle range, walking stride, walking frequency, walking speed, and walking duration are different, and different judgment conditions can be selected correspondingly. For example, in the case of a large walking stride, it can be judged whether an abnormality occurs according to the rate of change of the motion state data, and in the case of a small walking stride, it can be judged whether an abnormality occurs according to whether the motion state data exceeds a preset threshold . Specifically, it can be set or combined according to actual needs, which is not limited in this application.

In one embodiment, the anomaly detection strategy may include an anomaly detection model and anomaly judgment conditions. Specifically, the processor 106 may select characteristic variables from the training data of the training object and input them into the determined abnormality detection model, obtain an abnormality evaluation value after calculation by the abnormality detection model, and combine the abnormality evaluation value with the determined abnormality judgment Conditions are compared, and the final abnormality detection result is obtained according to the comparison. For example, when the abnormality evaluation value meets the abnormality judgment condition, it can be determined that the training object is currently abnormal.

Wherein, different anomaly detection levels may correspond to different anomaly detection strategies. Different anomaly detection levels can only correspond to different anomaly detection models, and there is no restriction on whether the abnormality judgment conditions are different; different anomaly detection levels can also only correspond to different anomaly judgment conditions, and there is no restriction on whether the anomaly detection models are different; of course , different anomaly detection levels can also correspond to different anomaly detection models and anomaly judgment conditions at the same time. Can be designed according to actual needs.

In one embodiment, the method further includes: acquiring data selection rules; selecting data from training data according to the data selection rules as abnormality detection data; judging whether the training object is abnormal according to the determined abnormality detection strategy and abnormality detection data.

In this embodiment, the processor 106 may also select part of the data from the training data set formed by the training data of the training object as the abnormality detection data for abnormality detection. More specifically, the processor 106 can determine the length of the time window for data collection according to the data screening rules, and use the time window to extend the length forward from the current time point in the training data set, and use the time window in the time window The training data is used as anomaly detection data.

In one embodiment, the data selection rule is to select according to at least one of anomaly detection level, training parameters and object parameters. Among them, the object parameters may include one or more of parameters such as age, height, weight, and athletic ability, and the training parameters may include joint movement angles, walking strides, walking strides, walking paces, and walking durations generated during training. One or more of the training related parameters.

Exemplarily, the length of the time window for data collection may be determined according to the abnormality detection level of the training object and the pace of the training object. Wherein, the length of the time window can be inversely proportional to the level of abnormal detection, that is, the higher the level of abnormal detection, the shorter the length of the time window, which can increase the detection frequency of abnormal detection and improve the sensitivity of abnormal judgment; the length of the time window It is inversely proportional to the current pace of the training object, that is, the faster the current pace, the shorter the length of the time window, so as to ensure that the collected data corresponds to a consistent motion state.

In an embodiment, the method further includes: judging that an abnormality occurs to the training object, and controlling the rehabilitation training equipment to start protection measures. Optionally, the convulsion protection measures include at least one of emergency stop, deceleration, switching training mode, warning and reversal.

In this embodiment, when an abnormality of the training object is detected, the processor 106 can control the rehabilitation training equipment to start protection measures, and timely activation of the protection measures can prevent the training equipment from causing secondary damage to the training object.

Further, different protection measures can be set for different anomaly detection levels. By setting different protection measures corresponding to different abnormality detection levels, targeted protection schemes can be given for different training objects, thereby improving the accuracy of abnormality protection in rehabilitation training.

In one embodiment, there is a corresponding relationship between the abnormality detection level and at least one training mode of the rehabilitation training equipment. At least one of joint motion angle range, motion stride and motion pace may be different between different training modes.

In this embodiment, a suitable training mode can be provided to the training object for the selected abnormality detection level, and different abnormality detection levels can be matched for the selected training mode for the user to choose. In this way, the selection operation steps of the user can be further reduced, and the risk of selection errors can be avoided.

Exemplarily, the processor 106 may control the rehabilitation training device to automatically adjust the training mode according to the acquired abnormality detection level, so that the rehabilitation training device adjusts to the training mode corresponding to the abnormality detection level.

Exemplarily, the processor 106 may also control the display to display one or more training patterns corresponding to the abnormality detection level on the display interface according to the acquired abnormality detection level, and the user may select one or more training modes displayed on the display. A selection is made among the training modes, and the processor adjusts according to the training mode selected by the user in response to the selection operation by the user through the display interface. The processor 106 can also match one or more abnormal detection levels according to the acquired training mode, and control the display to display one or more abnormal detection levels corresponding to the training mode on the display interface, and the user can select one or more abnormal detection levels displayed on the display. Select from a plurality of anomaly detection levels, and the processor 106 performs anomaly detection according to the anomaly detection level selected by the user in response to the user's selection operation through the display interface.

In one embodiment, the method further includes: dynamically adjusting the anomaly detection level during the training process in response to the level adjustment instruction.

In this embodiment, the processor 106 can also respond to the level adjustment instruction in real time and dynamically during the training process, and dynamically adjust the abnormality detection level, so that the abnormality detection strategy can be dynamically adjusted as the training progresses.

Exemplarily, during the training process, the processor 106 may receive a level adjustment instruction triggered by the user, and adjust the current anomaly detection level to the anomaly detection level specified by the user according to the level adjustment instruction.

Exemplarily, during the training process, after the level adjustment instruction is triggered, the processor 106 can intelligently analyze the training data of the training object in real time or periodically, and automatically match the corresponding abnormality detection level according to the data analysis result. Further, The processor 106 can control the abnormality detection level to automatically adjust from the current abnormality detection level to the abnormality detection level matching the data analysis result.

Below, combined with an application example, the anomaly detection method involved in this application will be further described in detail:

1. Get the anomaly detection level input by the user.

2. Obtain the training data of the training object to generate a training data set.

3. Obtain the data selection rules of the training data. The length of the time window for data analysis is determined according to one or more of the currently acquired anomaly detection level, training parameters and/or object parameters.

4. Determine the abnormality detection model according to the obtained abnormality detection level. Select the training data in the time window of the training data set, such as gait phase, joint angle, joint angular velocity and joint interaction torque, and combine the reference joint torque of the training object, abnormal detection level, etc., and use the above data as feature variable input to determine The anomaly detection model of , based on the functional correlation between the above feature variables, calculates the abnormal evaluation value of the training object.

5. According to the acquired abnormality detection level, determine an abnormality judgment condition, for example, determine an abnormality judgment threshold.

6. Compare the abnormality evaluation value with the abnormality judgment threshold, and when the abnormality evaluation value is greater than the abnormality judgment threshold, generate an abnormality detection result, and judge that the training object is abnormal.

7. When it is determined that the training object is abnormal, an alarm message is generated, and further, the rehabilitation training equipment can be controlled to start corresponding protection measures.

8. Receive the level adjustment instruction, and dynamically adjust the abnormality detection level according to the level adjustment instruction.

9. Adjust the training mode of the rehabilitation training equipment according to the adjusted abnormality detection level.

10. Repeat steps 2 to 9 above.

In one embodiment, as shown in FIG. 3 , an anomaly detection device is provided, which includes: a level acquisition module 310, a data acquisition module 320, a policy determination module 330 and an anomaly judgment module 340, wherein:

A level acquisition module 310, configured to acquire an abnormality detection level;

Data acquisition module 320, used to acquire the training data of the training object;

A strategy determination module 330, configured to determine a corresponding abnormality detection strategy according to the abnormality detection level;

The abnormality judging module 340 is configured to judge whether the training object is abnormal according to the determined abnormality detection strategy and the training data.

In one embodiment, the data acquisition module 320 is also used to acquire data selection rules, and selects data from the training data according to the data selection rules as abnormality detection data; the abnormality judgment module 340 is also used to determine the abnormality detection strategy according to the As well as the abnormality detection data, it is judged whether the training object is abnormal.

In one embodiment, the abnormality judging module 340 is further configured to judge that the training object is abnormal, and control the rehabilitation training equipment to start protection measures.

In one embodiment, the level acquisition module 310 is further configured to dynamically adjust the abnormality detection level during the training process in response to the level adjustment instruction.

For the specific limitation of the abnormality warning device, please refer to the above-mentioned limitation of the abnormality warning method, which will not be repeated here. Each module in the above-mentioned abnormal alarm device can be fully or partially realized by software, hardware and combinations thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, and can also be stored in the memory of the computer device in the form of software, so that the processor can invoke and execute the corresponding operations of the above-mentioned modules.

In one embodiment, the present application also provides a lower limb rehabilitation training device, as shown in FIG. 4 , the lower limb rehabilitation training device 400 may include:

The training component 402 and the abnormality detection system 404, wherein the training component 402 is used to support the training object and drive the training object to perform rehabilitation training. The anomaly detection system 404 can be built into the training component 402, or can be externally connected to the training component.

For the specific limitations of the abnormality detection method performed by the processor in the abnormality detection system 404 in the lower limb rehabilitation training device, please refer to the specific limitations of each step in the abnormality detection method above, which will not be repeated here.

In one embodiment, as shown in Fig. 5, Fig. 5 is a schematic structural diagram of the training components of the lower limb rehabilitation training device in one embodiment. In this embodiment, the training components may include a device hip joint 502 , a device knee joint 504 , and a device ankle joint 506 . In this embodiment, the training data of the training object may be acquired through the sensor 500 . The sensor 500 may be disposed on or near the location of at least one of the device hip joint 502 , the device knee joint 504 and the device ankle joint 506 .

In one embodiment, a computer device is provided, and the computer device may be a terminal device. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected through a system bus. Wherein, the processor of the computer device is used to provide calculation and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal via a network connection. The computer program implements the anomaly detection method when executed by the processor. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covered on the display screen, or a button, a trackball or a touch pad provided on the casing of the computer device , and can also be an external keyboard, touchpad, or mouse.

In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored on the memory and operable on the processor. When the processor executes the computer program, the following steps are implemented: obtaining an abnormality detection level; obtaining The training data of the training object; determine the corresponding abnormality detection strategy according to the abnormality detection level; judge whether the training object is abnormal according to the determined abnormality detection strategy and the training data.

In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented: obtaining the abnormality detection level; obtaining training data of the training object; according to the abnormality detection level Determine the corresponding anomaly detection strategy; judge whether the training object is abnormal according to the determined anomaly detection strategy and the training data.

Those of ordinary skill in the art can understand that the implementation of all or part of the processes in the above embodiments can be completed by instructing related hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium In this case, when the computer program is executed, it may include the processes of the embodiments of the above-mentioned methods. Wherein, any references to memory, storage, database or other media used in the various embodiments provided in the present application may include non-volatile and/or volatile memory. Nonvolatile memory can include read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Chain Synchlink DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, all possible combinations of the technical features in the above embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, they should be It is considered to be within the range described in this specification.

The above-mentioned embodiments only represent several implementation modes of the present application, and the description thereof is relatively specific and detailed, but it should not be construed as limiting the scope of the patent for the invention. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the scope of protection of the patent application should be based on the appended claims.

Claims (11)

1. An abnormality detection method applied to rehabilitation training equipment for providing lower limb rehabilitation training for a training subject, characterized by comprising the following steps:

obtaining an abnormality detection grade;

acquiring training data of a training object;

determining a corresponding abnormality detection strategy according to the abnormality detection grade;

and judging whether the training object is abnormal or not according to the determined abnormality detection strategy and the training data.

2. The method according to claim 1, wherein the anomaly detection strategy comprises an anomaly detection model and/or an anomaly determination condition.

3. The method according to claim 2, wherein the anomaly detection model comprises an evaluation feature and/or an evaluation relationship.

4. The method according to claim 1, wherein the method further comprises:

acquiring a data selection rule;

selecting data from the training data according to a data selection rule as abnormality detection data;

and judging whether the training object is abnormal or not according to the determined abnormality detection strategy and the abnormality detection data.

5. The method of claim 4, wherein the data selection rule is a data selection based on at least one of the anomaly detection level, a training parameter, and an object parameter.

6. The method according to any one of claims 1 to 5, further comprising:

and judging that the training object is abnormal, and controlling the rehabilitation training device to start protective measures.

7. The method of claim 6, wherein the protective measures include at least one of emergency stop, deceleration, switching training modes, alerting, and reversing.

8. The method of any one of claims 1 to 5, wherein the anomaly detection level corresponds to at least one training mode of the rehabilitation training device.

9. The method according to any one of claims 1 to 5, further comprising:

the anomaly detection level is dynamically adjusted during training in response to a level adjustment instruction.

10. An anomaly detection system for use in a rehabilitation training device for providing rehabilitation training of a lower limb to a training subject, comprising:

the grade acquisition component is used for acquiring an abnormality detection grade;

the data acquisition component is used for acquiring training data of a training object;

a processor coupled to the rank acquisition component and the data acquisition component, the processor configured to perform the method of any one of claims 1 to 9.

11. A lower limb rehabilitation training device, comprising: the anomaly detection system of claim 10.

Images