CN115494397A

CN115494397A - UPS state evaluation method and device based on BP neural network and storage medium

Info

Publication number: CN115494397A
Application number: CN202211120631.8A
Authority: CN
Inventors: 谷梦勖; 万衡; 唐许良
Original assignee: Shanghai Institute of Technology
Current assignee: Shanghai Institute of Technology
Priority date: 2022-09-15
Filing date: 2022-09-15
Publication date: 2022-12-20

Abstract

The invention provides a UPS state evaluation method based on a BP neural network, which comprises the following steps: setting UPS configuration parameters, selecting data in a UPS data acquisition system, and acquiring a module evaluation set and a UPS real-time monitoring data set; collecting a UPS real-time monitoring data set and a module score set, constructing a total data set in an arithmetic manner, and dividing the total data set into a training set, a verification set and a test set required by the arithmetic; adopting a sigmoid function as an activation function, building a BP neural network model, and inputting a selected training set for training; establishing an evaluation index of the model, and setting an optimal parameter of the algorithm through verification; and detecting whether the established model meets the set precision requirement or not by adopting the test set, if not, repeatedly executing until the established BP neural network model meets the requirement, and outputting the prediction score value of the model. The invention enables the UPS evaluation to be objectively fit with the real-time change of the monitoring data, and has the function of predicting the equipment state of the UPS, namely evaluating the future time period according to the maintenance condition.

Description

UPS state evaluation method and device based on BP neural network and storage medium

Technical Field

The invention relates to the field of state evaluation of enterprise power equipment, in particular to a state evaluation method and device of a UPS and a storage medium.

Background

The UPS system is a power backup device connected between an input power source and a load, and is capable of providing a grid-interference-free, voltage-stabilized, and frequency-stabilized power supply to an important load in the event of an external power failure. Therefore, the operation state of the UPS equipment and the health degree evaluation of the equipment are very important for ensuring the safe operation of the whole power supply system.

The UPS state evaluation system applied in many occasions at present mainly utilizes inherent experience to set weight for predicting the score value by human subjectively, and is deficient in the aspect of processing and processing of data.

This prior art solution also presents the following problems when in use:

the calculation formula of the UPS equipment scoring is too fixed, the real-time data of the UPS equipment is not fully utilized, and the actual running state of a nonlinear and unstable complex system of the UPS equipment is not met;

2. the problem that the UPS equipment data cannot be fully and flexibly collected without an input interface module because the parameters are set by people and are too fixed;

therefore, improvements to the above problems are needed to solve the above technical problems.

Disclosure of Invention

The invention aims to provide a UPS state evaluation method based on a BP neural network, so as to solve the problems in the background technology.

In order to achieve the purpose, the invention provides the following technical scheme:

a UPS state evaluation method based on a BP neural network. The method comprises the following steps:

s1, setting UPS configuration parameters, selecting data in a UPS data acquisition system, and acquiring a module evaluation set and a UPS real-time monitoring data set;

s2, summarizing the UPS real-time monitoring data set and the module scoring set, constructing a total data set in an arithmetic manner, and dividing the total data set into a training set, a verification set and a test set;

s3, adopting a sigmoid function as an activation function, building a BP neural network model, and inputting the selected training set for training;

s4, establishing an evaluation index of the model, and setting the optimal parameters of the algorithm through verification;

and S5, detecting whether the established model meets the set precision requirement or not by adopting the test set, and if not, repeatedly executing S3 and S4 until the established BP neural network model meets the requirement to obtain a prediction score value.

The UPS data acquisition system evaluates internal relevance data of the UPS state system, and needs to meet a large amount of dynamic real-time data with a certain sampling frequency, such as performance index parameters automatically generated by normal operation equipment.

The module evaluation set is used for performing auxiliary reference evaluation data on the UPS state evaluation system, has low occurrence frequency or event driving type, is greatly influenced by artificial experience, and mainly comprises the following steps: environment, UPS and storage battery fault conditions, preventive tests, maintenance and overhaul and reconstruction.

The configuration parameters include sampling frequency or data weighting coefficients. The sampling frequency comprises an acquisition period and a sampling granularity; the relative proportion coefficient between the evaluation set of the data weight coefficient determination module and the UPS real-time monitoring data set can be obtained through empirical values and can be further refined through historical data analysis.

Optionally, the UPS status scoring system includes a UPS device statistics data interface, and sets up a UPS system configuration and a maintenance process on the interactive interface. The problem that the UPS equipment real-time data cannot be fully acquired due to the fact that parameters are set manually and the UPS equipment real-time data cannot be acquired due to the fact that the UPS equipment real-time data cannot be set manually is effectively solved.

Optionally, the UPS state evaluation system is applied to an enterprise, and data volatility caused by different time use conditions of the power equipment needs to be fully considered. The UPS real-time monitoring data set constructs a required data set through an arithmetic progression in a mode of taking a week as a period and taking a half hour as an interval.

Optionally, the UPS real-time monitoring data set is selected according to the judgment of the UPS state operation condition, and includes ABC three-phase voltage output by ac, ABC three-phase voltage output by the UPS and ABC three-phase voltage data of an ac bypass, and the UPS and battery operation state module is constructed to participate in UPS overall scoring.

Optionally, the data preprocessing is performed on the module evaluation sets through sampling frequency or data weight coefficients.

Optionally, the specific process required for building the BP neural network model includes:

firstly, calculating an input value g of the hidden layer, and then calculating an output value h of the hidden layer according to the value:

h＝f(g)

the input value m of the output layer and the final output value y of the output layer are calculated by the same principle:

y＝f(m)

calculating the general error d of the output layer:

d＝(o-y)·y(1-y)

calculating the generalized error e of the hidden layer:

connection weight v and threshold γ are trimmed:

v(N+1)＝v(N)+α·d·h

γ(N+1)＝γ(N)+α·d

and then trimming the connection weight w and the threshold value theta:

w(N+1)＝w(N)+e·β·α

θ(N+1)＝θ(N)+e·β

and repeating the iteration process until the value output by the BP neural network converges to the error within the allowable range or the iteration times meet the requirement.

In the formula, except the identified letters, x is an input value, w is a connection weight from an input layer to a hidden layer, theta is an output threshold of a hidden layer unit, o is an expected output value, and alpha is a learning rate.

Optionally, the BP neural network model has three layers, including:

adopting sigmoid function as activation function to control output value in smaller value area, the formula is as follows:

selecting 9 types of data reflecting the operation states of the UPS and the storage battery and 5 module scores of the module scores as neurons of an input layer, namely the number of the input layer nodes is 14;

the total score of the corresponding UPS state evaluation system at the corresponding time point is taken as the neuron of the output layer, namely the number of the output layer nodes is 1.

Optionally, the evaluation index of the constructed model includes:

wherein, MAE is the average absolute error, MSE is the mean square error, RMSE is the root mean square error, Y is the actual value, W is the evaluation formula constructed by the invention, Y' is the predicted value, n is the number of the tested data, and t is the iteration number required by the simulation operation.

Optionally, the constructed evaluation formula W includes:

the invention takes the integral of the product of time and mean square error as the measurement standard for setting the iterative training times. Although the iterative training times are set more, the error of the model prediction value is smaller, the application scenario of the invention is a UPS state evaluation system in actual engineering, the real-time performance of a data simulation experiment is very important, and two factors of the prediction precision of the system and the running time required by simulation need to be comprehensively considered.

Optionally, the set precision requirement includes:

the final precision requirement of the method is set to be 0.1%, namely the final precision requirement can be output when the error is less than or equal to 0.1%, otherwise, S3 and S4 are repeatedly executed for continuing training and verification. And until the established BP neural network model meets the requirements, displaying the prediction score value of the BP neural network model through a DSP output module interface.

Optionally, the embodiment of the present specification further provides an apparatus, which may be regarded as a specific physical implementation manner for the method embodiment of the present invention described above. Wherein the apparatus comprises:

a processor; and a memory storing computer executable programs that can set UPS configuration parameters according to UPS specific model parameters, the executable programs, when executed, causing the processor to perform any of the methods described above, and display the final result.

The device can also comprise a UPS equipment statistical data interface, a user interaction module or an output display module; and setting UPS configuration parameters through a user interaction module, and outputting the prediction score value by an output display module. The problem that the UPS equipment real-time data cannot be fully acquired due to the fact that parameters are set manually and the UPS equipment real-time data cannot be fully acquired due to the fact that no real-time input interface module exists in the prior UPS equipment real-time data acquisition system is effectively improved.

The present specification also provides a computer readable storage medium, wherein the computer readable storage medium stores one or more programs which, when executed by a processor, implement any of the above methods.

In various technical solutions provided in the embodiments of the present description, by obtaining relevant data of a UPS, inputting control parameters, and using a running state module as an object, a UPS state evaluation method based on a BP neural network selects an intelligent control algorithm instead of a manual evaluation process, so that a UPS real-time evaluation criterion follows real-time changes of monitoring data.

The invention constructs a UPS evaluation system based on BP neural network algorithm, which has UPS device statistical data interface, sets the system constitution and maintenance flow of UPS on the interactive interface, and displays the health degree score data of UPS. The UPS evaluation criterion can be more suitable for real-time change of monitoring data, the problem that the effect of a traditional prediction method in a nonlinear and non-stable complex system of the real-time state of UPS equipment is poor is solved, the prediction function that the UPS is maintained according to the conditions, namely the equipment state in the future time period is evaluated is further realized, and the practical engineering value is achieved.

Drawings

Fig. 1 is a flowchart of a UPS state evaluation method based on a BP neural network according to an embodiment of the present invention;

fig. 2 is a scoring structure diagram of a UPS status scoring system according to an embodiment of the present invention;

FIG. 3 is a schematic structural diagram of an apparatus provided in an embodiment of the present disclosure;

fig. 4 is a schematic diagram of a computer-readable medium provided in an embodiment of the present specification.

Detailed Description

In order to make the implementation objects, technical solutions and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention are described in detail below with reference to the accompanying drawings in the embodiments of the present invention.

The data which influence equipment scoring through the historical data analysis of the UPS state scoring system comprise a large amount of dynamic real-time data and a data set which has low occurrence frequency or is event-driven and is influenced by artificial experience to a large extent; and different data have different importance for equipment evaluation.

Fig. 1 is a flowchart of a UPS state estimation method based on a BP neural network according to this embodiment, including the following steps:

s1: setting UPS configuration parameters, selecting data in a UPS data acquisition system, and acquiring a module evaluation set and a UPS real-time monitoring data set;

the constructed UPS state scoring system is provided with a UPS device statistical data interface, and related parameters such as the system composition, the maintenance process and the like of the UPS are set on an interactive interface. Therefore, the problems that the original UPS equipment is too fixed due to manual parameter setting and related data of the UPS equipment cannot be fully collected due to the fact that no real-time input interface module exists can be effectively solved.

The configuration parameters include sampling frequency or data weighting coefficients. The sampling frequency comprises an acquisition period and a sampling granularity; the data weight coefficient determining module is used for determining a proportionality coefficient between the evaluation set and the UPS real-time monitoring data set, and the proportionality coefficient can be obtained through empirical values and can be further refined through historical data analysis.

The module evaluation set is used for assisting reference evaluation data of the UPS state evaluation system, has low occurrence frequency or event driving type, is greatly influenced by artificial experience, and mainly comprises the following steps: environment, UPS and storage battery fault conditions, preventive tests, maintenance and overhaul and reconstruction. Can be independently determined by sampling frequency or data weight coefficient according to actual conditions.

The sampling frequency of the UPS real-time monitoring data set is not less than the sampling frequency of the module scoring set;

the operation of the UPS in the current state can be generally divided into three cases: the method comprises the steps of exiting operation, degrading operation and normally operating, and taking 9 types of data of ABC three-phase voltage output by alternating current, ABC three-phase voltage output by the UPS and ABC three-phase voltage of an alternating current bypass as objects adopted in the UPS data acquisition system according to the basis for judging the operation mode of the system.

S2: collecting a UPS real-time monitoring data set and a module score set, constructing a total data set in an arithmetic manner, and dividing the total data set into a training set, a verification set and a test set required by the arithmetic;

as shown in fig. 2, part of the real-time monitoring data of the UPS is selected as a component of an operation state module in the system, and the real-time monitoring data and the modules of the module evaluation set are collected to participate in the operation of the intelligent algorithm, so that the evaluation system is objective and has real-time performance.

Because the research object of the invention is a UPS state evaluation system applied in an enterprise, the data fluctuation caused by the use conditions of power equipment at different time needs to be fully considered, the invention constructs a required data set by an arithmetic progression in a way of taking a week as a period and taking half an hour as an interval, and the required data set is used as a grading summary of the UPS and the storage battery operation state and each module of the module grading set and is divided into a training set, a verification set and a test set according to the algorithm requirement to operate in a special DSP system.

S3: adopting sigmoid function as activation function, building BP neural network model, inputting selected training set for training, the specific flow of the algorithm is as follows:

since the output value can be controlled in a relatively small numerical region when the activation function in the BP neural network adopts an S-shaped nonlinear differentiable function, the invention adopts a sigmoid function, and the formula is as follows:

h＝f(g)

y＝f(m)

calculating the general error d of the output layer:

d＝(o-y)·y(1-y)

calculating the generalized error e of the hidden layer:

and (3) trimming connection weight v and threshold value gamma:

v(N+1)＝v(N)+α·d·h

γ(N+1)＝γ(N)+α·d

and then trimming the connection weight w and the threshold value theta:

w(N+1)＝w(N)+e·β·α

θ(N+1)＝θ(N)+e·β

In the formula, except for the marked letters, x is an input value, w is a connection weight from an input layer to a hidden layer, theta is an output threshold of a hidden layer unit, o is an expected output value, and alpha is a learning rate.

The invention selects the 9 types of data for judging the current state in the UPS real-time monitoring data and the scores of the modules in the module evaluation set as the neurons of the input layer, and the total scores of the modules corresponding to the UPS state evaluation system at the corresponding time points are used as the neurons of the output layer, thereby building a three-layer BP neural network model.

S4: and (4) establishing an evaluation index of the model, and setting the optimal parameters of the algorithm through verification. The setting of relevant parameters in the BP neural network model is of great importance to the influence of the overall simulation performance, and the parameter adjustment and evaluation of a verification data set through verification are required. The invention mainly evaluates the performance of the model by the following indexes:

wherein MAE is the average absolute error, MSE is the mean square error, RMSE is the root mean square error, Y is the actual value, Y' is the predicted value, and n is the number of data tested.

Meanwhile, although the iterative training times are set more, the error of the model prediction value is smaller, because the application scene of the invention is a UPS state evaluation system in practical engineering, the real-time performance of a data simulation experiment is very important, and two factors of the prediction precision of the system and the operation time required by simulation need to be comprehensively considered, the integral of the product of time and mean square error is used as the measurement standard for setting the iterative training times, and the formula W constructed by the invention is as follows:

where t is the number of iterations required for the simulation run.

S5: and detecting whether the established model meets the set precision requirement or not by adopting the test set, if not, repeatedly executing the step S3 and the step S4 until the established BP neural network model meets the requirement, and displaying the predicted score value of the BP neural network model through a DSP output module interface.

The final precision requirement set by the invention is 0.1 percent, namely the final precision requirement can be output when the error is less than or equal to 0.1 percent, otherwise, the S3 and the S4 are repeatedly executed for continuing training and verification until the established BP neural network model meets the requirement. Tests show that after test sample data are input, the predicted score value displayed through the DSP output module interface is high in coincidence degree with an actual value, the model precision can reach 99.97%, the change trends are the same, error change fluctuation is small, model prediction is stable, the model is established by applying a BP neural network algorithm to predict the system state score, and the method for refining the UPS evaluation criterion through real-time monitoring data and achieving system state score prediction is feasible.

Based on the same inventive concept, the embodiment of the specification further provides a device.

In the following, embodiments of the apparatus of the invention are described, which may be regarded as concrete physical embodiments for the above-described embodiments of the method of the invention. The details described in the device embodiments of the invention should be regarded as complementary to the above-described method embodiments; reference is made to the above-described method embodiments for details not disclosed in the apparatus embodiments of the invention.

Fig. 3 is a schematic structural diagram of an apparatus provided in an embodiment of the present disclosure. An apparatus E according to this embodiment of the invention is described below with reference to fig. 3. The apparatus E shown in fig. 3 is only an example and should not bring any limitation to the function and the scope of use of the embodiment of the present invention.

As shown in fig. 3, the apparatus E is in the form of a general purpose computing device. Components of the apparatus may include, but are not limited to: at least one processing unit E1, at least one memory unit E2, a bus E3 connecting different system components (including the memory unit E1 and the processing unit E2), a display unit E4, etc.

Wherein the storage unit stores program code executable by the processing unit E1 to cause the processing unit E1 to perform the steps according to various exemplary embodiments of the present invention described in the above-mentioned processing method section of the present specification. For example, the processing unit E1 may perform the steps as shown in fig. 1.

The storage unit E2 may include a readable medium in the form of a volatile memory unit, such as a random access memory unit (RAM) E21 and/or a cache memory unit E22, and may further include a read only memory unit (ROM) E23.

The storage unit E2 may also include a program/utility E24 having a set (at least one) of program modules E25, such program modules E25 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which, or some combination thereof, may comprise an implementation of a network environment.

Bus E3 may be a local bus representing one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, and so forth using any of a variety of bus architectures.

The apparatus E may also communicate with one or more external devices I (e.g., keyboard, pointing device, bluetooth device, etc.), with one or more devices that enable a user to interact with the apparatus E, and/or with any devices (e.g., router, modem, etc.) that enable the apparatus E to communicate with one or more other computing devices. This communication may be via an input/output (I/O) interface E5. Also, the device E may communicate with one or more networks (e.g., a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network, such as the internet) through the network adapter E6. The network adapter E6 can communicate with the other modules of the device E via the bus E3. It should be understood that although not shown in fig. 3, other hardware and/or software modules may be used in conjunction with apparatus E, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, among others.

Through the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments of the present invention described herein may be implemented by software, or by software in combination with necessary hardware. Therefore, the technical solution according to the embodiment of the present invention can be embodied in the form of a software product, which can be stored in a computer-readable storage medium (which can be a CD-ROM, a usb disk, a removable hard disk, etc.) or on a network, and includes several instructions to make a computing device (which can be a personal computer, a server, or a network device, etc.) execute the above-mentioned method according to the present invention. The computer program, when executed by a data processing apparatus, enables the computer readable medium to implement the above-described method of the invention, namely: such as the method shown in fig. 1.

The device provided by the embodiment can also comprise a UPS equipment statistical data interface, a user interaction module or an output display module; and setting UPS configuration parameters through a user interaction module, and outputting the prediction score value by an output display module. The problem that the UPS equipment real-time data cannot be fully acquired due to the fact that parameters are set manually and the UPS equipment real-time data cannot be fully acquired due to the fact that no real-time input interface module exists in the prior UPS equipment real-time data acquisition system is effectively improved.

A computer program implementing the method shown in fig. 1 may be stored on one or more computer readable media. The computer readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

The computer readable storage medium may include a propagated data signal with readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A readable storage medium may also be any readable medium that is not a readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a readable storage medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

Program code for carrying out operations for aspects of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device, or entirely on the remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computing device (e.g., through the internet using an internet service provider).

In summary, the invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art will appreciate that some or all of the functionality of some or all of the components in embodiments in accordance with the invention may be implemented in practice using a general purpose data processing device such as a microprocessor or a Digital Signal Processor (DSP). The present invention may also be embodied as apparatus or device programs (e.g., computer programs and computer program products) for performing a portion or all of the methods described herein. Such programs implementing the present invention may be stored on computer-readable media or may be in the form of one or more signals. Such a signal may be downloaded from an internet website or provided on a carrier signal or in any other form.

While the foregoing embodiments have described the objects, aspects and advantages of the present invention in further detail, it should be understood that the present invention is not inherently related to any particular computer or virtual machine, and that various general-purpose machines may implement the present invention. The invention is not to be considered as limited to the specific embodiments thereof, but is to be understood as being modified in all respects, all changes and equivalents that come within the spirit and scope of the invention.

The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments.

The above-mentioned embodiments are only used to illustrate the technical solution of the present invention, and do not limit the same. Any person skilled in the art may modify the above-described solutions and substitute part or all of the features without departing from the scope of the invention.

Claims

1. A UPS state evaluation method based on a BP neural network is characterized by comprising the following steps:

s4, establishing an evaluation index of the model, and setting an optimal parameter of the algorithm through verification;

2. The BP neural network-based UPS status assessment method of claim 1, wherein the module score set comprises: the method comprises the following steps of environment, UPS and storage battery fault conditions, preventive tests, maintenance and overhaul reconstruction; and carrying out data preprocessing on the module scores according to the sampling frequency or the data weight coefficient to obtain the required module score set.

3. The UPS state evaluation method based on the BP neural network according to claim 1, wherein the UPS monitors the data set in real time to judge the UPS state operation condition according to the selection, and the UPS and storage battery operation state module is constructed to participate in the UPS total scoring, wherein the selection comprises three-phase voltage of AC output, three-phase voltage of UPS output and three-phase voltage of AC bypass.

4. The UPS state estimation method according to claim 1, wherein the overall data set is composed of a module evaluation set and a UPS real-time monitoring data set, and the collection frequency is a period of one week and an interval of half an hour.

5. The method according to claim 1, wherein the specific process required for building the BP neural network model comprises:

h＝f(g)

y＝f(m)

calculating the general error d of the output layer:

d＝(o-y)·y(1-y)

calculating the generalized error e of the hidden layer:

connection weight v and threshold γ are trimmed:

v(N+1)＝v(N)+α·d·h

γ(N+1)＝γ(N)+α·d

and then trimming the connection weight w and the threshold value theta:

w(N+1)＝w(N)+e·β·α

θ(N+1)＝θ(N)+e·β

repeating the above process until the numerical value output by the BP neural network converges to the error in the allowable range, or the iteration times meet the requirements;

in the formula, x is an input value, w is a connection weight value from an input layer to a hidden layer, theta is an output threshold value of a hidden layer unit, o is an expected output value, and alpha is a learning rate.

6. The UPS state evaluation method based on the BP neural network according to claim 1, wherein the BP neural network model is 3 layers, 9 types of data and module scores reflecting the UPS and storage battery running states are selected as 5 module scores to serve as neurons of an input layer, the number of nodes of the input layer is 14, and the number of nodes of an output layer is 1.

7. The BP neural network-based UPS status assessment method of claim 1, wherein the set accuracy requirement is 0.1%.

8. An apparatus, characterized in that the apparatus comprises: a memory, a processor and a UPS state evaluation program stored on the memory and operable on the processor to run a BP neural network, the UPS state evaluation method program of the BP neural network, when executed by the processor, implementing the UPS state evaluation method based on the BP neural network according to any one of claims 1 to 7.

9. The device of claim 8, comprising a user interaction module or an output display module; and setting UPS configuration parameters through a user interaction module, and outputting the prediction score value by an output display module.

10. A computer-readable storage medium, wherein a BP neural network-based UPS state estimation method program is stored on the computer-readable storage medium, and when the BP neural network-based UPS state estimation method program is executed by a processor, the BP neural network-based UPS state estimation method according to any one of claims 1 to 7 is implemented.