A method and system for monitoring the state of an electrostatic servo mechanism based on a neural network

Abstract

The present invention discloses a method and system for monitoring the state of an electrostatic servo mechanism based on a neural network. The method comprises: establishing a deep neural network model; collecting the actuator speed, hydraulic pump inlet pressure, outlet pressure, hydraulic cylinder inlet pressure, outlet pressure, motor current, motor speed, boost tank pressure, and hydraulic oil temperature corresponding to the electrostatic servo mechanism under various displacement instructions in a normal state; using the motor speed as output data and the rest as input data, and training the deep neural network model to obtain a motor speed prediction model; using the motor current as output data and the rest as input data, and training to obtain a motor current prediction model; using the two prediction models to construct an electrostatic servo mechanism state monitoring index and obtain a fault threshold; obtaining the state data of the electrostatic servo mechanism to be monitored under specified displacement and speed instructions, calculating the monitoring index and comparing it with the threshold, and determining the state change of the electrostatic servo mechanism.

Description

Technical Field

[0001] The present invention belongs to the technical field of neural network, and in particular to a method and system for monitoring the state of an electrostatic pressure servo mechanism based on a neural network. Background Art

[0002] Electro-Hydrostatic Actuator (EHA), also known as electro-hydrostatic actuator, electro-hydrostatic servo system, is characterized by the use of "bidirectional quantitative hydraulic pump + hydraulic actuator" as a reducer and actuator, and the servo motor controls the output flow of the quantitative pump. Electro-hydrostatic servo mechanism takes into account the advantages of high efficiency and energy saving of electromechanical servo mechanism, easy use and maintenance, and high reliability and heavy load capacity of electro-hydraulic servo mechanism, and is widely used in aircraft, launch vehicles and other fields.

[0003] The most prominent requirement for servo mechanisms in aircraft and launch vehicles is high reliability. Therefore, it is an urgent issue to monitor the status of the electrostatic servo mechanism and ensure high reliability in executing tasks. In the prior art, when detecting whether there is an abnormality in the electrostatic servo mechanism, a single input displacement signal is usually used to obtain an output signal, and then the output signal and the value of the expected output signal are subjected to error analysis. If the error is within an acceptable range, it is determined that there is no abnormality in the electrostatic servo mechanism. However, when this method is used to detect the electrostatic servo mechanism, it can only determine whether the electrostatic servo mechanism can work, and it is impossible to monitor the degradation trend of the status of the electrostatic servo mechanism, resulting in a lack of basis for predicting the remaining life of the electrostatic servo mechanism. Summary of the invention

[0004] The purpose of the embodiments of the present application is to provide a method and system for monitoring the state of an electrostatic servo mechanism based on a neural network, aiming to solve the problem of how to monitor the state degradation process of the electrostatic servo mechanism and meet the needs of predicting the remaining life of the servo mechanism.

[0005] According to a first aspect of an embodiment of the present application, a method for monitoring the state of an electrostatic servo mechanism based on a neural network is provided, comprising:

[0006] Step (1): Establish a deep neural network model;

[0007] Step (2): Establish an electrostatic servo mechanism state database: collect the actuator speed x2, hydraulic pump inlet pressure x3, outlet pressure x4, hydraulic cylinder inlet pressure x5, outlet pressure x6, motor current x7, motor speed x8, boost tank pressure x9, hydraulic oil temperature x1, etc. corresponding to the electrostatic servo mechanism in normal state when it moves under various displacement x1 instructions. 10 data;

[0008] Step (3): using the motor speed x8 in each set of data in the electrostatic servo mechanism state database as the first output data and the remaining data as the first input data, training the deep neural network model to obtain a motor speed prediction model;

[0009] Step (4): using the motor current x7 in each group of data in the electrostatic servo mechanism state database as the second output data and the remaining data as the second input data, training the deep neural network model to obtain a motor current prediction model;

[0010] Step (5): Using the motor speed prediction model and the motor current prediction model, construct the electrostatic servo mechanism state monitoring index and obtain the fault threshold HI min ;

[0011] Step (6): Obtain the state data of the electrostatic servo mechanism to be monitored under the specified displacement and speed instructions, calculate the monitoring index HI, and compare the monitoring index HI with the threshold HI min Compare and determine the state change of the electrostatic servo mechanism.

[0012] Furthermore, the deep neural network model is a feedforward neural network model, including an input layer, a hidden layer and an output layer.

[0013] Furthermore, the step (3) comprises:

[0014] The electrostatic servo mechanism state database Divide into n groups according to time sequence, where i = 1, 2, ... N, n is an integer, N / n ≥ 100;

[0015] For each of the n time series data groups, the test, training, and validation data sets are allocated according to the percentages of r1, r2, and r3, where r1+r2+r3=100%;

[0016] Aggregate the obtained n test, training, and validation data sets respectively to form a first deep neural network model training data set;

[0017] Set the motor speed As the first output data, the remaining data is used as the first input data, and the deep neural network model is trained to obtain a motor speed prediction model.

[0018] Further, r1≥60%, r2≥20%, r2≥10%, r1+r2+r3=100%.

[0019] Furthermore, the jth test set data number is

[0020]

[0021] The training set data of group j is

[0022]

[0023] The j-th group validation dataset is

[0024]

[0025] Where j = 1, 2,…n.

[0026] Furthermore, the step (5) comprises:

[0027] (5.1) When the motor moves under the given actuator displacement x1 command, the displacement is collected. speed Hydraulic pump inlet pressure Outlet pressure Hydraulic cylinder inlet pressure Outlet pressure Motor current Motor speed Boost tank pressure Hydraulic oil temperature Time series data, where i = 1, 2, 3...N, N is the data sequence number;

[0028] (5.2) Using displacement speed Hydraulic pump inlet pressure Outlet pressure Hydraulic cylinder inlet pressure Outlet pressure Motor speed current The motor speed prediction model in step (3) is used to obtain the motor speed Using displacement speed Hydraulic pump inlet pressure Outlet pressure Hydraulic cylinder inlet pressure Outlet pressure Motor speed The motor current is obtained by the motor current prediction model described in step (4). The calculation method of the electrostatic servo mechanism status monitoring index HI is as follows:

[0029]

[0030] Among them, HI a is the average relative deviation of current, HI b is the average relative deviation of speed;

[0031] (5.3) By collecting historical data and artificially simulating various faults of the electrostatic servo mechanism, a database of the electrostatic servo mechanism under various fault conditions is obtained, and the range of the electrostatic servo mechanism state monitoring index HI is calculated according to the calculation method of step (5.2) [HI min ,I max ]; According to the calculation results, take the minimum value HI min As the threshold for determining electrostatic servo mechanism failure.

[0032] According to a second aspect of an embodiment of the present application, there is provided a neural network-based electrostatic servo mechanism state monitoring system, comprising:

[0033] The data acquisition unit is used to collect the corresponding actuator speed x2, hydraulic pump inlet pressure x3, outlet pressure x4, hydraulic cylinder inlet pressure x5, outlet pressure x6, motor current x7, motor speed x8, boost tank pressure x9, hydraulic oil temperature x1 when the electrostatic servo mechanism moves under various displacement x1 instructions in normal state. 10 data;

[0034] A data processing unit is used to establish a deep neural network model, and the motor speed x8 in each group of data in the electrostatic pressure servo mechanism state database is used as the first output data, and the remaining data is used as the first input data, and the deep neural network model is trained to obtain a motor speed prediction model; the motor current x7 in each group of data in the electrostatic pressure servo mechanism state database is used as the second output data, and the remaining data is used as the second input data, and the deep neural network model is trained to obtain a motor current prediction model;

[0035] The state monitoring and early warning unit is used to construct the state monitoring index of the electrostatic servo mechanism by using the motor speed prediction model and the motor current prediction model, and obtain the fault threshold HI min ; Obtain the state data of the electrostatic servo mechanism to be monitored under the specified displacement and speed instructions, calculate the monitoring index HI, and compare the monitoring index HI with the threshold HI min Compare and determine the state change of the electrostatic pressure servo mechanism;

[0036] The data storage module is used to store the state monitoring data of the electrostatic servo mechanism acquired by the data acquisition unit and the analysis results acquired by the state monitoring and early warning unit.

[0037] According to a third aspect of an embodiment of the present application, there is provided an electronic device, including:

[0038] one or more processors;

[0039] A memory for storing one or more programs;

[0040] When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in the first aspect.

[0041] According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, on which computer instructions are stored, and when the instructions are executed by a processor, the steps of the method described in the first aspect are implemented.

[0042] The technical solution provided by the embodiments of the present application may have the following beneficial effects:

[0043] It can be seen from the above embodiments that the present application solves the technical problem of monitoring the state degradation process of the electrostatic servo mechanism by adopting data prediction based on a neural network, and can realize state monitoring without adding sensors and data acquisition equipment.

[0044] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0046] Figure 1 The present invention is a flow chart showing a method for monitoring the state of an electrostatic servo mechanism based on a neural network according to an exemplary embodiment.

[0047] Figure 2 It is a block diagram of a state monitoring system of an electrostatic servo mechanism based on a neural network according to an exemplary embodiment.

[0048] Figure 3 The diagram is a schematic diagram of an electronic device according to an exemplary embodiment. DETAILED DESCRIPTION

[0049] Here, exemplary embodiments are described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application.

[0050] The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0051] It should be understood that although the terms first, second, third, etc. may be used in the present application to describe various information, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0052] Figure 1 is a flow chart of a method for monitoring the state of an electrostatic servo mechanism based on a neural network according to an exemplary embodiment. Figure 1 As shown, the method may include the following steps:

[0053] Step (1): Establish a deep neural network model;

[0054] Step (2): Establish an electrostatic servo mechanism state database: collect the corresponding hydraulic pump inlet pressure x3, outlet pressure x4, hydraulic cylinder inlet pressure x5, outlet pressure x6, motor current x7, motor speed x8, boost tank pressure x9, hydraulic oil temperature x1, and ...4, hydraulic cylinder inlet pressure x5, outlet pressure x6, hydraulic cylinder inlet pressure x5, hydraulic cylinder inlet pressure x6, hydraulic cylinder inlet pressure x7, hydraulic cylinder inlet pressure x8, hydraulic cylinder inlet pressure x9, hydraulic cylinder inlet pressure x1, hydraulic cylinder inlet pressure x1, hydraulic cylinder inlet pressure x2 10 data;

[0055] Step (3): using the motor speed x8 in each set of data in the electrostatic servo mechanism state database as the first output data and the remaining data as the first input data, training the deep neural network model to obtain a motor speed prediction model;

[0056] Step (4): using the motor current x7 in each group of data in the electrostatic servo mechanism state database as the second output data and the remaining data as the second input data, training the deep neural network model to obtain a motor current prediction model;

[0057] Step (5): Using the motor speed prediction model and the motor current prediction model, construct the electrostatic servo mechanism state monitoring index and obtain the fault threshold HI min ;

[0058] Step (6): Obtain the state data of the electrostatic servo mechanism to be monitored under the specified displacement and speed instructions, calculate the monitoring index HI, and compare the monitoring index HI with the threshold HI min Compare and determine the state change of the electrostatic servo mechanism.

[0059] It can be seen from the above embodiments that the present application solves the technical problem of monitoring the state degradation process of the electrostatic servo mechanism by adopting data prediction based on a neural network, and can realize state monitoring without adding sensors and data acquisition equipment.

[0060] Step (1): Establish a deep neural network model;

[0061] Specifically, a basic feedforward neural network model is established, including three parts: input layer, hidden layer and output layer. The feedforward neural network can theoretically approximate any continuous function and has strong multidimensional nonlinear data regression analysis capabilities. It can realize the actuator speed x2, hydraulic pump inlet pressure x3, hydraulic pump outlet pressure x4, hydraulic cylinder inlet pressure x5, outlet pressure x6, motor current x7, motor speed x8, boost tank pressure x9, hydraulic oil temperature x 10 Regression modeling of functional relationships between data, etc.

[0062] Step (2): Establish an electrostatic servo mechanism state database: collect the actuator speed x2, hydraulic pump inlet pressure x3, outlet pressure x4, hydraulic cylinder inlet pressure x5, outlet pressure x6, motor current x7, motor speed x8, boost tank pressure x9, hydraulic oil temperature x1, and the corresponding data when the electrostatic servo mechanism moves under different displacement x1 commands in the normal state. 10 data;

[0063] Specifically, when the electrostatic servo mechanism is in a normal state, the control runs under different displacement x1 programs, and the corresponding actuator speed x2, hydraulic pump inlet pressure x3, outlet pressure x4, hydraulic cylinder inlet pressure x5, outlet pressure x6, motor current x7, motor speed x8, boost tank pressure x9, hydraulic oil temperature x1 are collected. 10 Signal data, a total of data sets under normal conditions are collected N is the data sequence number.

[0064] Step (3): The motor speed of the electrostatic servo mechanism state database is as the first output data and the remaining data as the first input data, training the deep neural network model to obtain a motor speed prediction model;

[0065] Specifically, the measured N groups of data are first divided into n groups according to the time series, where n is an integer, N / n ≥ 100; then the test, training, and validation data sets are allocated to each of the n groups of time series data according to the percentages of r1, r2, and r3, where r1 ≥ 60%, r2 ≥ 20%, r2 ≥ 10%, and r1 + r2 + r3 = 100%;

[0066] The data number of the jth test set is

[0067]

[0068] The training set data of group j is

[0069]

[0070] The j-th group validation dataset is

[0071]

[0072] Finally, the obtained n test, training, and validation data sets are aggregated to form a deep neural network model training data set. As the first output data, where i=1, 2, ...N, and the remaining data are used as the first input data, the deep neural network model is trained to obtain a motor speed prediction model.

[0073] In one embodiment, N=100 sets of data are measured and divided into n=10 sets, and r1, r2 and r3 are 70%, 20% and 10% respectively.

[0074] Then the test set data numbers are: 1, 2, 3, 4, 5, 6, 7; 11, 12, ... 17; ... 91, 92, ... 97;

[0075] The training set data numbers are: 8,9; 18,19; ...; 98,99;

[0076] Then the validation set data numbers are: 10; 20; ...; 100;

[0077] It is best to merge the data according to the test set, training set and validation set data numbers in sequence to form a new data set.

[0078] When collecting state data of the electrostatic servo mechanism, the time series signal data under variable speed and variable load conditions is obtained. Reasonable allocation of test, training, and verification data sets is the key to improving the generalization ability of the motor speed prediction model. By allocating test, training, and verification data sets through the above technology, a more accurate prediction model can be obtained.

[0079] Step (4): The motor current in each set of data in the electrostatic servo mechanism state database is as the second output data and the remaining data as the second input data, training the deep neural network model to obtain a motor current prediction model;

[0080] Specifically, the measured N groups of data are first divided into n groups according to the time series, where n is an integer, N / n ≥ 100; then the test, training, and validation data sets are allocated to each of the n groups of time series data according to the percentages of r1, r2, and r3, where r1+r2+r3=100%; finally, the obtained n test, training, and validation data sets are aggregated to form a deep neural network model training data set. Finally, the motor speed is calculated. As the first output data, where i=1, 2, ...N, and the remaining data are used as the second input data, the deep neural network model is trained to obtain the motor current prediction model.

[0081] When collecting state data of the electrostatic servo mechanism, the time series signal data under variable speed and variable load conditions is obtained. Reasonable allocation of test, training, and verification data sets is the key to improving the generalization ability of the motor current prediction model. By allocating test, training, and verification data sets through the above technology, a more accurate prediction model can be obtained.

[0082] Step (5): Using the motor speed prediction model and the motor current prediction model, construct the electrostatic servo mechanism state monitoring index and obtain the fault threshold HI min ;

[0083] Specifically, step (5) may include the following sub-steps:

[0084] (5.1) Collect the displacement signal of the electrostatic servo mechanism during testing or operation Actuator speed Hydraulic pump inlet pressure Outlet pressure Hydraulic cylinder inlet pressure Outlet pressure Motor current Motor speed Boost tank pressure Hydraulic oil temperature data, where i = 1, 2, 3...N;

[0085] Specifically, when the electrostatic servo mechanism works according to the specified program, the displacement signal is collected. Actuator speed Hydraulic pump inlet pressure Outlet pressure Hydraulic cylinder inlet pressure Outlet pressure Motor current Motor speed Boost tank pressure Hydraulic oil temperature Signal data, forming time series data N is the data sequence number.

[0086] (5.2) Using the time series data obtained in (5.1) The motor speed prediction model in step (3) is used to obtain the motor speed under ideal conditions. The motor current prediction model in step (4) is used to obtain the motor current in the ideal state. Then, the electrostatic servo mechanism state monitoring index HI is calculated according to the method given in formula (1):

[0087]

[0088] Among them, HI a is the average relative deviation of current, HI b is the average relative deviation of speed;

[0089] Specifically, the time series data obtained using (5.1) Using displacement speed Hydraulic pump inlet pressure Outlet pressure Hydraulic cylinder inlet pressure Outlet pressure Motor speed current The motor speed prediction model in step (3) is used to obtain the motor speed under ideal conditions. Using displacement speed Hydraulic pump inlet pressure Outlet pressure Hydraulic cylinder inlet pressure Outlet pressure Motor speed The motor current prediction model in step (4) is used to obtain the motor current in the ideal state. Then, the electrostatic servo mechanism state monitoring index HI is calculated according to the method given in formula (1).

[0090] (5.3) Obtain a database of the electrostatic servo mechanism working under various fault conditions, and calculate the range of the electrostatic servo mechanism state monitoring index HI according to the calculation method of step (5.2) [HI min ,I max ]; According to the calculation results, take the minimum value HI min As the threshold for determining electrostatic servo mechanism failure.

[0091] Specifically, we first collect the signal data set of the electrostatic servo mechanism under fault condition when it runs according to the specified displacement program. Or artificially simulate various faults of the electrostatic servo mechanism and collect signal data sets under different fault conditions By collecting fault data through artificial simulation, fault data can be obtained before the electrostatic servo mechanism fails, thereby improving the accuracy of state detection and fault prediction.

[0092] Step (6): Obtain the state data of the electrostatic servo mechanism to be monitored under the specified displacement and speed instructions, calculate the monitoring index HI, and compare the monitoring index HI with the threshold HI min Compare and determine the state change of the electrostatic servo mechanism.

[0093] Specifically, the signal data under different fault states obtained in step (5) are used to calculate the monitoring index HI corresponding to different fault states, and then the minimum value HI is taken. min As a threshold value for determining the fault of the electrostatic pressure servo mechanism. By obtaining the fault threshold value through the above method, early warning of various electrostatic pressure servo mechanism faults can be achieved, thereby avoiding the occurrence of faults.

[0094] According to the above steps (1)-(6), the state data of EHA in normal state and fault states such as underpressure of booster tank, internal leakage, and oil filter blockage are measured respectively, including the corresponding hydraulic pump inlet pressure x3, outlet pressure x4, hydraulic cylinder inlet pressure x5, outlet pressure x6, motor current x7, motor (pump) speed x8, booster tank pressure x9, hydraulic oil temperature x1, and so on. 10 Data; then according to step (1) to step (5), the health indicators under different states are calculated, as shown in the table. The implementation case results show that when the EHA fails, the health state and fault degree of the electrostatic servo mechanism can be judged according to the method of the present invention.

[0095] Table 1 Comparison of HI under different fault conditions

[0096]

[0097] Corresponding to the aforementioned embodiment of the state monitoring method of the neural network-based electrostatic servo mechanism, the present application also provides an embodiment of the state monitoring system of the neural network-based electrostatic servo mechanism.

[0098] Figure 2 is a block diagram of a state monitoring system of an electrostatic servo mechanism based on a neural network according to an exemplary embodiment. Figure 2, the system may include:

[0099] The data acquisition unit 21 is used to collect the corresponding actuator speed x2, hydraulic pump inlet pressure x3, outlet pressure x4, hydraulic cylinder inlet pressure x5, outlet pressure x6, motor current x7, motor speed x8, boost tank pressure x9, hydraulic oil temperature x1, etc. when the electrostatic servo mechanism moves under various displacement x1 instructions in a normal state. 10 data;

[0100] The data processing unit 22 is used to establish a deep neural network model, and uses the motor speed x8 in each group of data in the electrostatic pressure servo mechanism state database as the first output data and the remaining data as the first input data to train the deep neural network model to obtain a motor speed prediction model; uses the motor current x7 in each group of data in the electrostatic pressure servo mechanism state database as the second output data and the remaining data as the second input data to train the deep neural network model to obtain a motor current prediction model;

[0101] The state monitoring and early warning unit 23 is used to construct the state monitoring index of the electrostatic servo mechanism by using the motor speed prediction model and the motor current prediction model, and obtain the fault threshold HI min ; Obtain the state data of the electrostatic servo mechanism to be monitored under the specified displacement and speed instructions, calculate the monitoring index HI, and compare the monitoring index HI with the threshold HI min Compare and determine the state change of the electrostatic pressure servo mechanism;

[0102] The data storage module 24 is used to store the state monitoring data and analysis results of the electrostatic servo mechanism.

[0103] Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0104] For the device embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The device embodiment described above is only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present application scheme. A person of ordinary skill in the art can understand and implement it without paying any creative work.

[0105] Accordingly, the present application also provides an electronic device, comprising: one or more processors; a memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the state monitoring method of the electrostatic servo mechanism based on the neural network as described above. Figure 3 As shown, a hardware structure diagram of a state monitoring method of an electrostatic servo mechanism based on a neural network provided by an embodiment of the present invention is provided for any device with data processing capability, except Figure 3 In addition to the processor, memory, DMA controller, disk, and non-volatile memory shown, any device with data processing capabilities in which the apparatus in the embodiment is located may also include other hardware, which will not be described in detail, generally based on the actual functions of the device with data processing capabilities.

[0106] Accordingly, the present application also provides a computer-readable storage medium on which computer instructions are stored, and when the instructions are executed by the processor, the state monitoring method of the electrostatic servo mechanism based on the neural network as described above is implemented. The computer-readable storage medium can be an internal storage unit of any device with data processing capabilities described in any of the aforementioned embodiments, such as a hard disk or a memory. The computer-readable storage medium can also be an external storage device of a wind turbine, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), an SD card, a flash card (Flash Card), etc. equipped on the device. Further, the computer-readable storage medium can also include both an internal storage unit of any device with data processing capabilities and an external storage device. The computer-readable storage medium is used to store the computer program and other programs and data required by any device with data processing capabilities, and can also be used to temporarily store data that has been output or is to be output.

[0107] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the contents disclosed herein. The present application is intended to cover any variations, uses or adaptations of the present application, which follow the general principles of the present application and include common knowledge or customary technical means in the art that are not disclosed in the present application.

[0108] It should be understood that the present application is not limited to the exact construction that has been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof.

Claims

1. A method for monitoring the state of an electrostatic servo mechanism based on a neural network, characterized in that: include: Step (1): Establish a deep neural network model; Step (2): Establish an electrostatic servo mechanism state database: collect the electrostatic servo mechanism in various displacement states under normal conditions The corresponding actuator speed when moving under command , Hydraulic pump inlet pressure , outlet pressure , hydraulic cylinder inlet pressure , outlet pressure , motor current , motor speed , boost tank pressure , hydraulic oil temperature data; Step (3): The motor speed in each set of data in the electrostatic servo mechanism state database is as the first output data and the remaining data as the first input data, training the deep neural network model to obtain a motor speed prediction model; Step (4): The motor current in each set of data in the electrostatic servo mechanism state database is as the second output data and the remaining data as the second input data, training the deep neural network model to obtain a motor current prediction model; Step (5): Using the motor speed prediction model and the motor current prediction model, construct the electrostatic servo mechanism state monitoring index and obtain the fault threshold ; Step (6): Obtain the state data of the electrostatic servo mechanism to be monitored under the specified displacement and speed instructions, calculate the monitoring index HI, and compare the monitoring index HI with the threshold value Compare and determine the state change of the electrostatic pressure servo mechanism; Wherein, the step (5) comprises: (5.1) Collect the motor displacement at a given actuator When commanding movement, collect displacement ,speed , hydraulic pump inlet pressure , outlet pressure , hydraulic cylinder inlet pressure , outlet pressure , motor current , motor speed , boost tank pressure , hydraulic oil temperature Time series data, where i=1,2,3...N, N is the data sequence number; (5.2) Using displacement ,speed , hydraulic pump inlet pressure , outlet pressure , hydraulic cylinder inlet pressure , outlet pressure , motor speed current , boost tank pressure , hydraulic oil temperature The motor speed prediction model described in step (3) is used to obtain the motor speed ; Using displacement ,speed , hydraulic pump inlet pressure , outlet pressure , hydraulic cylinder inlet pressure , outlet pressure , motor speed , boost tank pressure , hydraulic oil temperature The motor current is obtained by the motor current prediction model described in step (4). ; Then the calculation method of the electrostatic servo mechanism state monitoring index HI is as follows: （1） in, is the average relative deviation of current, is the average relative deviation of speed; (5.3) By collecting historical data and artificially simulating various faults of the electrostatic servo mechanism, a database of the electrostatic servo mechanism under various fault conditions is obtained, and the range of the electrostatic servo mechanism state monitoring index HI is calculated according to the calculation method of step (5.2) [ ]; According to the calculation results, take the minimum value As the threshold for determining electrostatic servo mechanism failure.

2. The method according to claim 1, characterized in that The deep neural network model is a feedforward neural network model, including an input layer, a hidden layer and an output layer.

3. The method according to claim 1, characterized in that The step (3) comprises: The electrostatic servo mechanism state database Divide into n groups according to time sequence, where i=1,2,…N, n is an integer, ; For each group of time series data, , and Percentage is used for test, training, and validation data set allocation, where ; Aggregate the obtained n test, training, and validation data sets respectively to form a first deep neural network model training data set; Set the motor speed As the first output data, the remaining data is used as the first input data, and the deep neural network model is trained to obtain a motor speed prediction model.

4. The method according to claim 3, characterized in that ， ， ， 。 5. The method according to claim 4, characterized in that The data number of the jth test set is ， The training set data of group j is ， The j-th group validation dataset is ， Where j=1,2,…n.

6. A system using the state monitoring method of the electrostatic servo mechanism based on a neural network as claimed in claim 1, characterized in that: include: The data acquisition unit is used to collect the data of the electrostatic servo mechanism under various displacement conditions in normal state. The corresponding actuator speed when moving under command , Hydraulic pump inlet pressure , outlet pressure , hydraulic cylinder inlet pressure , outlet pressure , motor current , motor speed , boost tank pressure , hydraulic oil temperature data; A data processing unit is used to establish a deep neural network model to convert the motor speed in each set of data in the electrostatic servo mechanism state database into as the first output data, and the remaining data as the first input data, to train the deep neural network model to obtain a motor speed prediction model; the motor current in each group of data in the electrostatic servo mechanism state database is as the second output data and the remaining data as the second input data, training the deep neural network model to obtain a motor current prediction model; The state monitoring and early warning unit is used to construct the state monitoring index of the electrostatic servo mechanism and obtain the fault threshold value by using the motor speed prediction model and the motor current prediction model. ; Obtain the state data of the electrostatic servo mechanism to be monitored under the specified displacement and speed instructions, calculate the monitoring index HI, and compare the monitoring index HI with the threshold Compare and determine the state change of the electrostatic pressure servo mechanism; The data storage module is used to store the state monitoring data of the electrostatic servo mechanism acquired by the data acquisition unit and the analysis results acquired by the state monitoring and early warning unit.

7. An electronic device, characterized in that: include: one or more processors; A memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having computer instructions stored thereon, characterized in that: When the instruction is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images