CN115577979A - A Method for RPA Energy Efficiency Evaluation in the Scenario of Batch New Installation of Low-Voltage Equipment - Google Patents

Abstract

The invention discloses a method for evaluating RPA energy efficiency under a scene of new batch installation of low-voltage equipment, which comprises the following steps: step A: collecting parameters of a system performance level and parameters of a system environment level in real time; and B, step B: establishing an environment parameter factor; step C: establishing an evaluation index system; step D: quantifying an evaluation index; step E: determining a pairwise judgment matrix and a fuzzy relation matrix; step F: and judging the system efficiency. On the theoretical basis of the efficiency evaluation technology, the invention provides an efficiency evaluation algorithm based on the environment fuzzy factor by combining the relevant technical basis and characteristics of the RPA. Under the background of RPA efficiency evaluation, the algorithm has better environmental adaptability than the traditional algorithm, can better evaluate the performance of the RPA, and optimizes the resource management efficiency of the RPA.

Description

A Method for RPA Energy Efficiency Evaluation in the Scenario of Batch New Installation of Low-Voltage Equipment

technical field

The invention belongs to the technical field of performance evaluation of RPA in the scene of batch new installation of low-voltage equipment in the power marketing industry, and specifically relates to a method for evaluating RPA energy efficiency in the scene of batch new installation of low-voltage equipment.

Background technique

RPA (Robotic Process Automation), also known as Digital Workforce (DWF), has the characteristics of anthropomorphism, non-intrusion, zero integration, and short cycle. It simulates human operations through computer programming or auxiliary software. The rules automatically execute process tasks, replace or assist humans to complete related computer operations, simplify business processes, eliminate bottlenecks in the process, complete repetitive and regular tasks, and bring more efficient working models. Improve work quality, reduce enterprise risk, effectively improve employee work efficiency and business service level, and ease the work burden at the grassroots level. Among them, the anthropomorphic feature refers to: by simulating the operation process of human and computer systems, it can replace human beings to complete process processing tasks. The non-invasive feature means that RPA runs at a higher software level, which does not invade and affect existing software systems, but operates these systems at the presentation level. The zero-integration feature means that RPA does not need to change the current system architecture, and uses technology to help enterprises improve their performance while maintaining the stable and reliable functions of the existing IT systems of the enterprise. The short cycle feature means that under normal circumstances, it only takes 2 to 3 days for a person familiar with the business process to develop and launch a moderately difficult RPA application. Compared with traditional development models (such as Java, etc.), the cycle of RPA is shortened by half.

The performance evaluation of traditional RPA refers to the health status and operation status evaluation of RPA, such as running time, running quantity, and resource load. At present, from the perspective of digital transformation, there is a lack of RPA performance evaluation strategies from a business perspective. Therefore, it is impossible to establish the presentation form, operation status, value connotation, and workload evaluation of RPA for the business. On this basis, how to accurately and effectively evaluate the performance of the RPA system, so as to optimize the performance of the RPA system has become an important issue.

The business scenario of batch new installation of low-voltage equipment for power marketing includes three business subsets: batch new installation of low-voltage public buildings in the community, batch information entry of low-voltage residents in the community, and splitting of work orders for batch new installation of low-voltage in the community. Among them, the batch new installation of low-voltage public buildings in the community includes the following steps: work order association, address maintenance, electricity consumption information filling, user classification, offer data import, fee control information filling, batch user information modification, zero-yuan prefabrication, etc. The batch information entry of low-voltage residents in the community includes the following steps: work order association, address maintenance, electricity consumption information filling, user classification, offer data import, judgment whether to split work orders, fee control information filling, batch user information modification, zero yuan prefabrication Wait. The division of low-voltage batch new installation work orders in the community includes the following steps: log in to the 186 system to read the work order to split the address, process the detailed address and unit address, query the work order, obtain the address of each page in the split page, address comparison and Tick addresses, send emails, etc.

In the marketing scenario of the electric power industry, the operating rules involved in the traditional low-voltage public building batch new installation business process are relatively fixed, but there are certain repetitive operations and heavy workload. For example, in the electricity price configuration of newly installed public construction meters in the community, it is necessary to manually configure the industry classification and electricity consumption category in the business acceptance process according to the electricity price rules, and the number of configurations is large. Manual configuration has long time-consuming, high personnel investment, and manual configuration errors. Disadvantages, resulting in reduced timeliness of customer service. Therefore, in order to effectively shorten the time limit for business processing and improve customer service levels, relying on the marketing SG186 system, deploy RPA tools and connect business rules, so as to realize the automation of manual operation procedures for business personnel, improve service efficiency, and enhance the accuracy of business operations. Efficiency refers to the effectiveness of achieving predetermined goals and the degree to which RPA, as a digital employee, can improve the value of quality and efficiency. Performance evaluation is based on the RPA platform and is business-oriented. It conducts quantitative and conclusive evaluations on the performance indicators such as the scale, quality, and execution status of RPA applications. In order to compare the performance changes before and after the application of RPA, it is necessary to provide a method for evaluating the energy efficiency of RPA in the scenario of batch new installation of low-voltage equipment.

Contents of the invention

The purpose of the present invention is to provide a method for evaluating RPA energy efficiency in the scene of batch new installation of low-voltage equipment.

For realizing the purpose of the present invention, the technical scheme provided by the present invention is as follows:

A method for evaluating RPA energy efficiency under the scene of batch new installation of low-voltage equipment, comprising the following steps:

Step A: Collect system performance level parameters and collect system environment level parameters in real time to form the initial parameters of the performance evaluation algorithm; wherein, the system performance level parameters include disk storage parameters, network parameters, and system computing power parameters . System expansion parameters, cache space parameters; wherein, the parameters at the system environment level include organization status parameters, business execution status parameters, and operation status parameters;

Step B: Establish environmental parameter factors, specifically as follows:

Step B-1: Environmental parameter selection

According to the environmental state of the RPA system, including 56 states, select the corresponding environmental fuzzy factor. The selected environmental fuzzy factor can represent the current RPA environmental state. Each environmental fuzzy factor represents different meanings when it participates in subsequent algorithm operations. ;

Step B-2: Confirm the corresponding environmental blur factor

Different fuzzy operators are determined according to different environmental conditions. When the network bandwidth is 10M, 20M, 50M, 100M, 200M, 500M, the corresponding fuzzy operators are {0, 0.2, 0.4, 0.6, 0.8, 1}, when When the memory usage rate is 10%, 20%, 40%, 60%, 80%, and 100%, the corresponding fuzzy operator is {0, 0.1, 0.3, 0.5, 0.7, 0.9, 1};

Step C: Establish an evaluation index system

Through the functional analysis and index research of the cognitive RPA system, three categories of performance evaluation factors are established,

The first category of evaluation factors is organizational status, which consists of RPA scale and RPA application effects, including the number of robots, robot/employee ratio, robot working hours, robot business domain distribution, and robot association;

The second type of evaluation factor is business execution status, which consists of system performance statistics, RPA application effects and RPA stability, including system load, process execution time, total cost ROI, process execution success rate, and manual maintenance times;

The third type of evaluation factor is the operation status. The operation status includes RPA application effect and RPA stability, including manual execution time, process execution time, average employee satisfaction, RPA coverage business ratio, and manual maintenance times;

Step D: Quantify Evaluation Metrics

Quantitatively calculate the various evaluation indicators in step C into the system, so as to perform subsequent calculations;

Among them, the number of robots refers to: For a single business domain, the more robots there are, the more people use digital tools, and the stronger the corresponding information acquisition and robot scheduling capabilities;

Robot/employee ratio means: for a single business domain, the larger the robot/employee ratio, the higher the proportion of repetitive work replaced by robots. This index can be directly calculated through the first-level index, which can reflect the scale of RPA;

The working hours of robots refer to: the working hours of a single process, the working hours of a single business domain, and the overall working hours;

The distribution of robot business domains refers to: the distribution of robot business domains. On the one hand, it can analyze the degree of digital transformation of various departments, and on the other hand, it can also serve as a reference for the business direction of RPA’s future development;

Robot relevance means: starting from the overall business of the group, there are business flows interacting with multiple departments, and robots communicate and cooperate with each other;

Step E: Establish pairwise judgment matrix and fuzzy relationship matrix. The specific fuzzy comprehensive evaluation method is as follows:

Step E-1: Build a set of evaluation indicators: U={u ₁ , u ₂ ,..., u _m };

Step E-2: Build an evaluation grade set: V={v ₁ , v ₂ ,...,v _n }; the number of evaluation grades is 5, which is {excellent, good, medium, poor, extremely poor};

Step E-3: Construct the fuzzy relationship matrix R=(r _ij ) _m×n , and evaluate each index element, where r _ij is the element u _i in the evaluation index set U corresponding to the element v _j in the evaluation level set V affiliation of

Step E-4: Construct the weight vector A _: each element of the weight vector A is the weight of the corresponding element in the evaluation index set U, where:

Step E-5: Combine R and A into the final result B. The final result B is calculated from the weight vector A and the fuzzy relationship matrix R. "." is some kind of fuzzy operator, the specific form is determined by the specific algorithm, among which:

Step E-6: Select the corresponding threshold according to the actual situation, and rate B according to the threshold, and the rating and scoring result is the performance evaluation result;

Input the judgment matrix according to the expert’s index comparison, and then pass the consistency check of the analytic hierarchy process to establish the weight of each index, and establish the judgment matrix for batch new installation of low-voltage public buildings in the marketing business scenario, as shown in formula 3;

The establishment of the fuzzy relationship matrix is strongly related to the aforementioned selected environmental background;

Step F: Judging System Performance

Among them, E _i represents the performance value of the i-th large index under the second level, W _ij represents the weight of the j-th index under the i-th index, and I _ij represents the quantized value of the j-th index under the i-th index Finally, from bottom to top, the top-level efficacy value can be calculated according to formula 4, and finally the result Q is judged according to the interval of the efficacy value. The division of the efficacy interval is shown in formula 5.

Wherein, in the step A, the disk storage parameters include the total disk space for deployment and the average disk usage.

Wherein, in the step A, the network parameters include the number of deployed physical nodes, the average network bandwidth usage, the number of interaction requests and the cloud service rate.

Wherein, in the step A, the parameters of the system computing power include the total amount of deployed CPU computing power and the average CPU load;

Wherein, in the step A, the cache space parameters include the total deployed memory and the average memory usage.

Wherein, in the step A, the system expansion aspect includes the usage of third-party AI components.

Wherein, in the step A, the organizational state parameters are composed of RPA scale and RPA application effect, including the number of robots, robot/employee ratio, robot working hours, robot business domain distribution, and robot association.

Wherein, in the step A, the business execution state parameters are composed of system performance statistics, RPA application effects and RPA stability, including system load, process execution time, total cost ROI, process execution success rate, and manual maintenance times .

Wherein, in the step A, the operation status parameters are composed of RPA application effects and RPA stability, including manual execution time, process execution time, average employee satisfaction, RPA coverage business ratio, and manual maintenance times.

Wherein, in the step D, the working time of the robot refers to: it is divided into single-process working time, single-business domain working time, and overall working time. The goal of digital efficiency improvement can only be achieved if the coefficient of single process duration/manpower duration is less than 1; the working hours of a single business domain is an indicator for evaluating the degree of digitization of each department and the effect of RPA on improving quality and efficiency, and is a collection of working hours of all processes in this business domain ; As a key indicator of the use of digital employees in the business as a whole, the overall working hours are the sum of the working hours of all business domains. The longer the overall working hours, the higher the utilization rate of robots.

Compared with the prior art, the beneficial effects of the present invention are as follows:

The present invention proposes an efficiency evaluation algorithm based on environmental fuzzy factors based on the theoretical basis of the efficiency evaluation technology, and then combining the relevant technical foundation and characteristics of RPA. In the context of RPA performance evaluation, this algorithm can have better environmental adaptability than traditional algorithms, can better evaluate the performance of RPA, and optimize the resource management efficiency of RPA.

Description of drawings

Fig. 1 is a schematic flow sheet of the method in the present invention;

Fig. 2 is a schematic diagram of the RPA system environment parameter system in the present invention;

Fig. 3 is a schematic diagram of the RPA index evaluation system in the present invention.

detailed description

The technical solutions in the embodiments of the present invention are clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

As shown in Figures 1-3, this embodiment involves a method for evaluating RPA energy efficiency in the scenario of batch new installation of low-voltage equipment, which specifically includes the following steps:

Step A: Data Collection Module

External input of specific system performance parameters: This step mainly receives some parameters of the system and obtains some environmental information of the current system environment in real time, which constitutes the initial parameters of the performance evaluation algorithm.

(1) Collect parameters at the system environment level

Accurately quantify the environmental parameters of the RPA running machine and input them into the data collection module. The disk storage includes the total disk space of the deployment and the average disk usage. The network aspect includes the number of deployed physical nodes, the average network bandwidth usage, the number of interactive requests, and the cloud service rate. The computing power of the system includes the total amount of memory deployed and the average memory usage. System expansion includes the usage of third-party AI components.

(2) Collect parameters at the system performance level

Accurately quantify the performance parameters of the RPA system and input them into the data collection module. Specifically, it includes organizational status, business execution status, and operational status. Among them, organizational status is composed of RPA scale and RPA application effect, mainly including the number of robots, robot/employee ratio, robot working hours, robot business domain distribution, robot association, etc. Business execution status consists of system performance statistics, RPA application effects, and RPA stability, mainly including system load, process execution time, total cost ROI, process execution success rate, and manual maintenance times. Operation status includes RPA application effect and RPA stability, mainly including manual execution time, process execution time, average employee satisfaction, RPA coverage business ratio, manual maintenance times, etc.

Step B: Establish environmental parameter factors

(1) Selection of environmental parameters

Environmental parameters The RPA system is in different environments, and its performance evaluation cannot be fully modeled. For example, under the premise of controlling variables, a network with a large bandwidth and a fast network is definitely better than a network with a small bandwidth and a slow network, which is directly determined by the network situation. After the traditional algorithm calculates the same performance value, it is unscientific and unreasonable to evaluate both good network conditions and poor network conditions as "good".

First, according to the environmental state of the RPA system, there are a total of 56 situations, and then the corresponding environmental fuzzy factor is selected. At this time, the selected environmental fuzzy factor represents the only one that can represent the current RPA environmental state. Each environmental fuzzy factor represents a different meaning when it participates in the subsequent algorithm operation. For example, when other parameters are the same, the RPA operation quality when the network is good is more likely to be excellent than the overall performance when the network is poor.

(2) Confirm the corresponding environmental fuzzy factor

Different fuzzy operators are determined according to different environmental conditions. When the network bandwidth is 10M, 20M, 50M, 100M, 200M, 500M, the corresponding fuzzy operators are {0, 0.2, 0.4, 0.6, 0.8, 1}, when When the memory usage rate is 10%, 20%, 40%, 60%, 80%, and 100%, the corresponding fuzzy operator is {0, 0.1, 0.3, 0.5, 0.7, 0.9, 1}.

Step C: Establish an evaluation index system

The effectiveness of the RPA system under cognitive power informatization comes directly from the final evaluation of various indicators of the system, so it is necessary to establish a scientific and reasonable evaluation system that meets business needs. Through the functional analysis and index investigation of the cognitive RPA system, three categories of performance evaluation factors were finally established, as shown in Figure 3. The first category of evaluation factors is organizational status, which is composed of RPA scale and RPA application effects, mainly including the number of robots, robot/employee ratio, working hours of robots, distribution of robot business domains, and robot associations. The second type of evaluation factor is business execution status, which is composed of system performance statistics, RPA application effect and RPA stability, mainly including system load, process execution time, total cost ROI, process execution success rate, and manual maintenance times Wait. The third type of evaluation factor is the operation status. The operation status includes the effect of RPA application and the stability of RPA, mainly including manual execution time, process execution time, average employee satisfaction, RPA coverage business ratio, and manual maintenance times.

The influencing factors of the cognitive RPA system are complex. How to mathematically model these influencing factors and convert various indicators into intuitive and quantitative numerical variables is the key task of evaluating the indicator system.

Step D: Quantify Evaluation Metrics

The indicators in Figure 3 need to be quantitatively calculated into the system for subsequent calculations. A brief description of the key indicators that affect performance, including the number of robots, robot/employee ratio, working hours of robots, distribution of robot business domains, and robot associations, etc.

(1) Number of robots: For a single business domain, the more robots there are, the more people use digital tools, and the stronger the corresponding information acquisition and robot scheduling capabilities. However, due to other factors such as the difficulty of business processes, the relationship between the number of robots and efficiency is not a simple linear relationship.

(2) Robot/employee ratio: For a single business domain, the larger the robot/employee ratio, the higher the proportion of repetitive work replaced by robots. This index can be directly calculated through the first-level index, which can reflect the scale of RPA from the side.

(3) Robot working hours: divided into single-process working hours, single-business domain working hours, and overall working hours. As a first-level indicator, the working time of a single process needs to be evaluated with the calculation of the working time of manual tasks. The goal of digital efficiency improvement can only be achieved by requiring the coefficient of single process time/manual time to be less than 1.

⑷The working time of a single business domain is an index to evaluate the degree of digitization of each department and the effect of RPA on improving quality and efficiency. It is a collection of working hours of all processes in this business domain.

(5) The overall working hours, as a key indicator of the use of digital employees in the business as a whole, is the sum of the working hours of all business domains. The longer the overall working hours, the higher the utilization rate of robots.

⑹ Distribution of robot business domains: The distribution of robot business domains can analyze the degree of digital transformation of various departments on the one hand, and can serve as a reference for the future development of RPA on the other hand.

⑺Robot relevance: Starting from the overall business of the group, there is a business flow of multi-department interaction. Robots communicate and cooperate with each other, avoiding the hysteresis of manual interaction, which is an important indicator of the overall application effect of RPA.

Step E: Establish pairwise judgment matrix and fuzzy relationship matrix

The specific fuzzy comprehensive evaluation method is as follows:

a) Build an evaluation index set: U={u ₁ , u ₂ ,..., u _m };

b) Constructing a set of evaluation levels: V={v ₁ , v ₂ , ..., v _n }; generally speaking, the number of evaluation levels should exceed four, but not more than five. {excellent, good, average, poor, very poor}.

c) Construct a fuzzy relationship matrix R=(r _ij ) _m×n , and evaluate each index element, where r _ij is the affiliation relationship of element u _i in the evaluation index set U corresponding to element v _j in the evaluation level set V .

d) Construction of weight vector A _: each element of weight vector A is the weight of the corresponding element in the evaluation index set U, where:

e) R and A are synthesized into the final result B. The final result B is calculated from the weight vector A and the fuzzy relationship matrix R. "." is some kind of fuzzy operator, the specific form is determined by the specific algorithm, among which:

f) Select the corresponding threshold according to the actual situation, and rate B according to the threshold, and the rating and scoring result is the performance evaluation result.

Finally, compare the input judgment matrix according to the indicators of experts, and then pass the consistency check of the analytic hierarchy process to establish the weight of each indicator. Finally, after extensive research and comprehensive comparison, the judgment matrix for batch new installation of low-voltage public buildings in the community in the marketing business scenario introduced in this article is established, as shown in Formula 3.

Since this application introduces the environmental fuzzy factor, the establishment of the fuzzy relationship matrix is strongly related to the aforementioned selected environmental background. So far, the pairwise judgment matrix and fuzzy relationship matrix are obtained, which is also the core step of the satellite effectiveness evaluation scheme based on environmental fuzzy factors.

Step F: Judging System Performance

Among them, E _i represents the performance value of the i-th large index under the second level, W _ij represents the weight of the j-th index under the i-th index, and I _ij represents the quantized value of the j-th index under the i-th index evaluation value. Finally, from bottom to top, the top-level performance value can be finally calculated according to formula 4. Finally, the judgment of the result Q is carried out according to the interval where the efficacy value is located. Wherein, the division of the efficiency interval is shown in formula 5.

It should be noted that in this application, the important implementation steps for batch new installation of low-voltage public buildings in the low-voltage side business community of electric power marketing are as follows: first, the RPA robot uses the application number of the community work order in the excel table to increase and change the capacity of the new installation in the system. Enter the query interface through the electricity/business expansion query/work order query path, enter the application number and click Query, click the application number of the work order in the work order list below, and then select the customer's natural information to obtain the customer number; then the RPA robot clicks on the queried work order The application number at the bottom of the list enters the work order query interface; secondly, the RPA robot opens the new installation capacity increase and changes electricity consumption/business acceptance/business acceptance/customer natural information, enters the customer number, presses Enter, selects the new installation work order of the community, and clicks Start Associated business, associate a low-voltage batch work order; finally, the RPA robot clicks OK after selecting low-voltage batch new installation for the associated business.

Specifically include the following:

S1: The RPA robot opens the browser webpage, enters the account password, and logs in to the marketing system;

S2: The RPA robot reads the table information and determines which business subset to execute. If it is low-voltage residents and public buildings, execute S3. If it is low-voltage work order splitting, execute S13;

S3: The RPA robot enters the query interface according to the application number of the work order of the community, and enters the query interface according to the new installation and capacity increase and change of electricity consumption/business expansion query/work order query path in the system. After entering the application number, click Query, and click the work order in the work order list below After the application number, select the customer natural information to get the customer number. Then click the application number at the bottom of the queried work order list to enter the work order query interface, and then open the new installation capacity increase and change power consumption/business acceptance/business acceptance/customer natural information, enter the customer number, click Enter, and select the new installation in the community Work order, click to start associated business, associate a low-voltage batch work order, and finally select the associated business as low-voltage public construction batch or low-voltage residential batch according to the judgment result of S2, and then click OK;

S4: Judging address maintenance, if the address in the excel table cannot be found and selected in the system, it is judged that the address does not exist, and an address needs to be added, and S5 is executed; if there is data in the address part of the form, fill in the address in the address , there is no need to add an address, execute S6;

S5: New installation, capacity increase and change of power consumption - Auxiliary management - Function - Standard address code maintenance - Add, select Tianjin as the superior address, select the category to be added in the category (street, road, neighborhood committee, community), and click on the name Enter the name of the street, road, neighborhood committee, and community, and click Save;

S6: In the electricity application information interface, the robot operation business subcategory selects new installations in batches for residents or new installations for public buildings according to the types of residents and public buildings. The robot fills in the electricity consumption address, contact person, mobile phone, application method, power supply unit, urban and rural category, contact type, electricity consumption category, supply voltage, industry classification, user classification, electricity bill type, and transfer sign according to the information in the excel sheet. Fill in the electricity address according to whether there is data in the excel sheet. If there is no data, the address does not need to be filled in. If the upper-level address is filled in, the lower-level address will disappear. It needs to be obtained and judged. If it disappears, it needs to be filled in again;

S7: Read the offer folder provided by the business personnel, open the offer data import interface, select low-voltage batch new offer import, input the low-voltage batch new installation application number, click Browse to identify and upload the data folder in the offer folder "Batch User Information.xml" to "Residential Building Details", select "Overwrite Import" in the drop-down option box, and click to start importing;

S8: The RPA robot judges whether the work order is split, and the robot queries the excel table "whether there is a household with an area greater than 180 square meters that needs to be equipped with a three-phase meter" and "whether there is a corridor light that configures residential electricity", if "Yes" Then split the work orders for households with a user area greater than 180 square meters and households with "corridor lights" and "stairway lights" (with an area of 0) in their electricity addresses (for example, "Whether there are households with an area greater than 180 square meters need to configure three If you select "Yes" in the field of phase table and the area of all users in the robot identification batch information is greater than 180, then there is no need to split the work order, and directly operate the entire work order according to the work order with an area greater than 180). The robot needs to perform the work order splitting operation according to the following steps: the robot clicks on the new installation to increase capacity and change power consumption-auxiliary management-function-work order splitting, and the robot enters the application number of the low-voltage batch work order to be split in the application number , the robot clicks the Enter key, selects the residential building information that appears below, and clicks to split the worksheet by each household;

S9: The RPA robot fills in the cost control information, enters the cost control information interface, and the robot reads the original work order in the excel table, the work order with an area greater than 180 and the cost control method of the corridor light work order. If you choose Energy Meter Fee Control: Select All, choose Prepaid Payment Type, choose Energy Meter Fee Control as Fee Control Method, click to select all application options, and click Save. Residents implement S10, and public buildings implement S11;

S10: The RPA robot performs batch modification and preservation of batch user assignment information and batch user information;

S11: Batch user information: Click Batch Edit, choose electricity consumption category and industry classification according to the table below, choose power institution counter charge as payment method, choose low-voltage non-residential user category, choose meter reading settlement for electricity bill settlement method, and choose 380V for voltage level , the electricity bill notification method is selected by the robot according to the excel table;

S12: Determine whether zero-yuan prefabrication: The robot reads the original work order in the excel table, the work order with an area greater than 180 and the corridor light work order are zero-yuan prefabrication. There are three options:

① All users: Select Yes on the "Whether Zero Prefabrication" interface, select "Yes", select all users in the user list on the left, select all users for the application scope, click Save, and click Apply.

②No: No need to operate this interface.

③ Check the user on the page: Select Yes in the "Whether Zero Prefabrication" interface, and the robot reads the details uploaded by the business personnel that need to apply zero prefabrication (Format: XX Unit House Number of XX Building in XX Community No.) and unit, the robot automatically identifies and checks the user list according to the application details and units on the left side of the page, checks the user on the application range selection page, clicks Save, and clicks Apply.

S13: The RPA robot reads offline data, and the robot reads the execl form downloaded from the platform and stores the details and unit addresses that need to be split, and processes the data;

S14: The robot queries the work order number that needs to be split, and the work order splitting operation is performed according to the following steps: the robot clicks on new installation, capacity increase and change power consumption-auxiliary management-function-work order splitting, and the robot enters the required work order at the application number. The application number of the split low-voltage batch work order, the robot clicks the Enter key, selects the residential building information that appears below, and clicks to split the work order by each household. According to the address details and unit numbers in the execl table that need to be split, the robot automatically recognizes all communities and addresses that need to be split, and ticks the check box on the left to split;

S15: After splitting, record the split work order number and save it in excel;

(X：长度像素数；Y：宽度像素数；Z：屏幕尺寸即对角线长度)。根据两者的相对位置，可以计算需要设置的偏移量A、B。Wherein, in the step S3, the target position is calculated through the position of the anchor point by means of resolution offset. Anchor position (X ₁ , Y ₁ ), target position (X ₂ , Y ₂ ), resolution PPI. The resolution is calculated as

(X: length in pixels; Y: width in pixels; Z: screen size or diagonal length). According to the relative position of the two, the offsets A and B that need to be set can be calculated.

X ₂ =X ₁ +A

Y ₂ =Y ₁ +B

Wherein, in the step S8, the weight standardization method is adopted, and the business experts score the residential user area, the floor of the community, the corridor lights and the number of users, etc., and calculate and judge the work order splitting situation according to the scoring weight combined with CRITIC .

The process of calculating the split work order is as follows:

CRITIC takes into account the correlation between indicators. The larger the standard deviation, the greater the fluctuation, that is, the greater the value gap between the various programs, the higher the weight will be; the conflict between indicators is represented by the correlation coefficient. If There is a strong positive correlation between the two indicators, indicating that the smaller the conflict, the lower the weight will be.

Normalize each factor according to the quantity of each option as:

For positive indicators:

For negative indicators:

Metric Variability:

Expressed in the form of standard deviation, S _j represents the standard deviation of the jth index

Index Conflict:

Expressed in the form of correlation coefficient, r _ij represents the correlation coefficient between evaluation indicators i and j

Information volume:

The larger C _j is, the greater the role of the jth evaluation index in the entire evaluation index system is, and more weight should be assigned to it

Weights:

根据上述权重计算方式，依据小区各指标数据可以得出小区工单的拆分情况，将工单拆分成1至3个工单，然后分别对各工单进行单独处理。To derive an objective weight for each metric

According to the above-mentioned weight calculation method, the division of work orders in the district can be obtained according to the data of each indicator in the district, and the work order is divided into 1 to 3 work orders, and then each work order is processed separately.

Use RPA to implement the batch new installation process of low-voltage equipment. In order to compare the performance changes before and after the use of RPA, a method of RPA performance evaluation is established. Performance refers to the effectiveness of achieving the predetermined goals, and the degree to which RPA as a digital employee can improve the value of quality and efficiency. Performance evaluation is based on the RPA platform and is business-oriented. It conducts quantitative and conclusive evaluations on the performance indicators such as the scale, quality, and execution status of RPA applications.

Finally, it should be noted that the above-mentioned embodiments are only used to illustrate and describe the present invention, and are not intended to limit the present invention to the scope of the described embodiments. In addition, those skilled in the art can understand that the present invention is not limited to the above-mentioned embodiments, and more variations and modifications can be made according to the teachings of the present invention, and these variations and modifications all fall within the scope of the present invention. .

Claims (10)

1. A method for evaluating RPA energy efficiency under a scene of new batch installation of low-voltage equipment is characterized by comprising the following steps:

step A: collecting parameters of a system performance level and parameters of a system environment level in real time to form initial parameters of an efficiency evaluation algorithm; the parameters of the system performance level comprise parameters in disk storage, parameters in network, parameters in system calculation, parameters in system expansion and parameters in cache space; the parameters of the system environment level comprise an organization state parameter, a service execution state parameter and an operation state parameter;

and B: establishing an environment parameter factor, specifically comprising the following steps:

step B-1: environmental parameter selection

Selecting corresponding environment fuzzy factors according to the environment states of the RPA system, including 56 states, wherein the selected environment fuzzy factors can represent the current RPA environment state, and the meanings of the environment fuzzy factors are different when the environment fuzzy factors participate in the subsequent algorithm operation;

step B-2: identifying corresponding ambient ambiguity factors

Determining different fuzzy operators according to different environment states, wherein when the network bandwidth is 10M, 20M, 50M, 100M, 200M and 500M, the corresponding fuzzy operator is {0,0.2,0.4,0.6,0.8 and 1}, and when the memory utilization rate is 10%, 20%, 40%, 60%, 80% and 100%, the corresponding fuzzy operator is {0,0.1,0.3,0.5,0.7,0.9 and 1};

and C: establishing an evaluation index system

Through the functional analysis and index investigation of the cognitive RPA system, three types of efficiency evaluation factors are established,

the first type of assessment factor is an organization state, wherein the organization state consists of RPA scale and RPA application effect, and comprises the number of robots, robot/employee ratio, long working time of the robots, robot service domain distribution and robot association;

the second type of evaluation factors are service execution states, and the service execution states comprise system performance statistics, RPA application effects and RPA stability, and comprise system load conditions, flow execution time, cost ROI total amount, flow execution success rate and manual maintenance times;

the third type of evaluation factors are operation states, wherein the operation states comprise RPA application effects and RPA stability, and comprise manual execution time, process execution time, average satisfaction of staff, RPA coverage service proportion and manual maintenance times;

step D: quantitative evaluation index

C, quantitatively calculating each evaluation index in the step C into a system, and performing subsequent operation;

wherein, the number of robots means: for a single service domain, the more the number of robots is, the more the number of people using digital tools is, and the stronger the corresponding information acquisition and robot scheduling capabilities are;

robot/employee ratio refers to: for a single service domain, the larger the robot/worker ratio is, the higher the ratio of replacing the repetitive work by the robot is, the index can be directly calculated through a first-level index, and the size of the RPA scale can be laterally reflected;

the long working time of the robot means that: the method comprises the following steps of dividing the working time of a single flow, the working time of a single service domain and the total working time;

the robot service domain distribution means that: the distribution situation of the robot service domain can analyze the digitalized transformation degree of each department on one hand, and can also serve as a reference for the future service direction of RPA development on the other hand;

robot relevance refers to: starting from the group overall service, a multi-department interactive service flow exists, and robots are in mutual communication and cooperation;

step E: determining a pairwise judgment matrix and a fuzzy relation matrix, wherein the specific fuzzy comprehensive judgment method comprises the following steps:

step E-1: constructing an evaluation index set: u = { U = ₁ ,u ₂ ,…,u _m }；

Step E-2: and (3) constructing an evaluation grade set: v = { V = ₁ ,v ₂ ,…,v _n }; the number of evaluation grades is 5, and is { excellent, good, medium, poor, extremely poor };

step E-3: constructing a fuzzy relation matrix R = (R) _ij ) _m×n And evaluating each index element, wherein r _ij For element U in evaluation index set U _i Corresponding to the element V in the evaluation level set V _j Membership of (d);

step E-4: constructing a weight vector A: each element of the weight vector a is a weight of a corresponding element in the evaluation index set U, where:

step E-5: r and a are synthesized into a final result B, which is calculated from the weight vector a and the fuzzy relation matrix R, in equation 2 ". "is a fuzzy operator, the specific form is determined by specific algorithm, wherein:

step E-6: selecting a corresponding threshold value according to the actual condition, and grading and scoring B according to the threshold value, wherein the grading and scoring result is an efficiency evaluation result;

comparing and inputting a judgment matrix according to indexes of experts, and then verifying consistency through an analytic hierarchy process so as to establish each index weight and establish a judgment matrix for newly installing a cell in a low-voltage public building batch in a marketing business scene as shown in a formula 3;

establishing a fuzzy relation matrix which is strongly related to the selected environment background;

step F: judging system efficiency

Wherein, E _i Represents the efficacy value of the ith large index at the second level, W _ij Represents the weight of the jth index under the ith index, I _ij And finally, calculating to obtain a top efficiency value according to a formula 4 by representing the evaluation value after quantization of the jth index under the ith index from bottom to top, and judging a result Q according to the interval of the efficiency value, wherein the division of the efficiency interval is shown as a formula 5:

。

2. the method for RPA energy efficiency assessment under the scenario of new batch installation of low-voltage devices as claimed in claim 1, wherein in step A, the parameters in terms of disk storage include total disk space deployment and average disk usage rate.

3. The method for RPA energy efficiency assessment under the scenario of new batch installation of low-voltage devices according to claim 1, wherein in step A, the network-aspect parameters include a number of deployed physical nodes, an average network bandwidth usage amount, a number of interactive requests, and a cloud service rate.

4. The method for RPA energy efficiency assessment according to claim 1, wherein in step A, the system computing power parameters include total amount of CPU computing power and average CPU load.

5. The method according to claim 1, wherein in step a, the cache space parameters include total memory deployment amount and average memory usage rate.

6. The method for evaluating the RPA energy efficiency under the low-voltage equipment new batch installation scene as claimed in claim 1, wherein in the step A, the system expansion aspect includes third-party AI component usage.

7. The method for RPA energy efficiency assessment under the scenario of new batch installation of low-voltage devices as claimed in claim 1, wherein in step A, the organization state parameters are composed of RPA scale and RPA application effect, including number of robots, robot/employee ratio, robot working time, robot service domain distribution, and robot association.

8. The method for RPA energy efficiency assessment under the scenario of new batch installation of low-voltage devices according to claim 1, wherein in step A, the service execution state parameters include system performance statistics, RPA application effects, and RPA stability, including system load conditions, process execution time, cost ROI total, process execution success rate, and manual maintenance times.

9. The method for RPA energy efficiency assessment under the scenario of new batch installation of low-voltage devices according to claim 1, wherein in step A, the operation state parameters include RPA application effect and RPA stability, including manual execution time, process execution time, average employee satisfaction, RPA coverage service proportion, and manual maintenance times.

10. The method for evaluating the RPA energy efficiency under the new low-voltage equipment batch installation scene according to claim 1, wherein in step D, the operating time of the robot is: the method comprises the following steps of dividing the method into single-flow working time, single-service domain working time and total working time, wherein the single-flow working time is used as a first-level index and needs to be calculated and evaluated with the manual task working time, and the digital efficiency improvement target can be achieved only when the single-flow working time/manual working time requirement coefficient is smaller than 1; the working duration of a single service domain is an index for evaluating the digitization degree and the RPA quality and efficiency improvement effect of each department, and is a collection of the working durations of all processes of the service domain; the total working duration is used as a key index of the operation condition of the digital staff in the whole business, and is the sum of the working durations of all business domains, and the longer the total working duration is, the higher the utilization rate of the robot is.

Images