Performance prediction method, device and equipment of electric actuating mechanism and medium

Abstract

The invention relates to an electric actuating mechanism performance prediction method and device, equipment and a medium, and the method comprises the steps: obtaining a current state parameter of an electric actuating mechanism, inputting the current state parameter of a transmission part into a transmission part performance prediction model, and obtaining a performance prediction result of the transmission part at a future moment; inputting the current state parameters of the circuit board into the circuit board performance prediction model to obtain a performance prediction result of the circuit board at the future moment, and determining a performance prediction result of the electric actuating mechanism at the future moment according to the performance prediction result of the transmission part at the future moment and the performance prediction result of the circuit board at the future moment; according to the method, the performance degradation condition of the electric actuating mechanism can be predicted in time, maintenance personnel can conveniently evaluate possible faults of the electric actuating mechanism in time, the maintenance personnel are assisted in performing maintenance service in time, and the conditions that the electric actuating mechanism is shut down due to faults and normal process production is affected are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of electric actuator performance prediction, and in particular to an electric actuator performance prediction method, device, equipment and medium. Background Art

[0002] An electric actuator is a device that converts electrical energy into mechanical energy. It is composed of motors, electromagnetic actuators, controllers, reducers and other components. It automatically completes the adjustment action by receiving control signals from regulators, industrial computers, etc. It is an important part of industrial automation process control systems.

[0003] At present, the method for performance testing of electric actuators is to collect the real-time status parameters of the electric actuators and compare the real-time status parameters with the empirical threshold to determine whether the performance of the electric actuators has declined. For example, the movement trajectory of a normal and healthy electric actuator from a certain starting position to a certain end position usually has a certain stable, continuous and smooth curve trajectory, and the torque size during operation is also within the predetermined range and stable. When there is a fault in the action component of the electric actuator, the movement trajectory may not be smooth, the torque size may exceed the preset range, and the electric actuator may not reach the specified end position normally.

[0004] However, the performance test results of the electric actuator are obtained after the moving parts of the electric actuator have already had an obvious fault. Obviously, the above method has a lag in the fault detection results of the electric actuator. When the fault of the electric actuator is detected, it has already caused an adverse impact on the process production at the production site.

[0005] Therefore, it is necessary to improve the current performance testing method of electric actuators. Summary of the invention

[0006] In view of the above-mentioned shortcomings of the prior art, the present application provides a method, device, equipment and medium for predicting the performance of an electric actuator to solve the above-mentioned technical problems.

[0007] The present invention provides a method for predicting the performance of an electric actuator, the method comprising: obtaining current state parameters of the electric actuator; the electric actuator comprising a transmission component and a circuit board; inputting the current state parameters of the transmission component into a transmission component performance prediction model to obtain a performance prediction result of the transmission component at a future moment, the transmission component performance prediction model being obtained by training a preset transmission component performance prediction model with transmission component sample data; inputting the current state parameters of the circuit board into a circuit board performance prediction model to obtain a performance prediction result of the circuit board at a future moment; the circuit board performance prediction model being obtained by training a preset circuit board performance prediction model with circuit board sample data; and determining the performance prediction result of the electric actuator at a future moment according to the performance prediction result of the transmission component at a future moment and the performance prediction result of the circuit board at a future moment.

[0008] In one embodiment of the present invention, a preset transmission component performance prediction model is trained through transmission component sample data to obtain the transmission component performance prediction model, and the process includes: inputting the transmission component sample data into a preset feature extraction model to obtain frequency domain feature signals and time domain feature signals of the target transmission component at different historical moments; the transmission component sample data includes usage status parameters of the target transmission component at different historical moments; the model of the target transmission component is the same as the model of the transmission component; training a preset time series prediction model through the frequency domain feature signal to obtain a trained time series prediction model; inputting the frequency domain feature signal into the trained time series prediction model to obtain linear prediction results of the target transmission component at different historical moments; and based on the usage status parameters and the linear prediction results, determining the linear residuals of the target transmission component at different historical moments; and using the linear residuals to train a preset bidirectional long short-term memory neural network model The method comprises the steps of: performing training to obtain a trained bidirectional long short-term memory neural network model; performing nonlinear prediction on the frequency domain feature signal by using the trained bidirectional long short-term memory neural network model to obtain nonlinear prediction results of the target transmission component at different historical moments; using a preset fusion model to fuse the linear prediction result and the nonlinear prediction result to obtain fused prediction results of the target transmission component at different historical moments, and representing the fused prediction results by a target frequency domain signal; using a preset transformation model to transform the target frequency domain signal and the time domain feature signal to obtain performance prediction results of the target transmission component at different historical moments; based on the performance prediction results of the target transmission component at different historical moments, combining the preset feature extraction model, the trained time series prediction model, the trained bidirectional long short-term memory neural network model, the preset fusion model and the preset transformation model to obtain a performance prediction model of the transmission component.

[0009] In one embodiment of the present invention, the process of inputting the transmission component sample data into the preset feature extraction model to obtain the frequency domain feature signal and the time domain feature signal of the target transmission component at different historical moments includes: calculating the root mean square of the transmission component sample data, and using the root mean square of the transmission component sample data as the time domain feature signal of the target transmission component at different historical moments; performing empirical wavelet transform on the time domain feature signal to obtain a low-frequency sub-layer signal of the time domain feature signal, wherein the low frequency is less than a preset frequency threshold; performing normalization calculation on the low-frequency sub-layer signal of the time domain feature signal to obtain the frequency domain feature signal of the target transmission component at different historical moments.

[0010] In one embodiment of the present invention, the process of fusing the linear prediction result and the nonlinear prediction result to obtain the fused prediction result of the target transmission component at different historical moments includes: selecting a moment as the target moment from different historical moments, multiplying the linear prediction result of the target moment by a preset linear result weight to obtain a first product; multiplying the nonlinear prediction result of the target moment by a preset nonlinear result weight to obtain a second product; taking the sum of the first product and the second product as the fused prediction result of the target transmission component at the target moment; reselecting a moment from different historical moments, and calculating the fused prediction result of the target transmission component at the corresponding moment, until all historical moments are selected and the calculation is terminated to obtain the fused prediction results of the target transmission component at different historical moments.

[0011] In one embodiment of the present invention, the process of transforming the target frequency domain signal and the time domain characteristic signal to obtain the performance prediction results of the target transmission component at different historical moments includes: combining the target frequency domain signal and the time domain characteristic signal belonging to the same historical moment until the combination of the target frequency domain signals and the time domain characteristic signals at all historical moments is completed to obtain the combined characteristic signals of the target transmission component at different historical moments; performing an empirical wavelet inverse transform on the combined characteristic signal to obtain the performance prediction results of the target transmission component at different historical moments.

[0012] In one embodiment of the present invention, the process of determining the performance prediction result of the electric actuator at a future moment according to the performance prediction result of the transmission component at a future moment and the performance prediction result of the circuit board at a future moment includes: multiplying the performance prediction result of the transmission component at a future moment by a preset transmission component result weight to obtain a third product; multiplying the performance prediction result of the circuit board at a future moment by a preset circuit board result weight to obtain a fourth product; and taking the sum of the third product and the fourth product as the performance prediction result of the electric actuator at a future moment.

[0013] In one embodiment of the present invention, before training a preset transmission component performance prediction model through transmission component sample data to obtain the transmission component performance prediction model, the method further includes: controlling the motor of the actuator prototype to act under a preset test environment, and, in the process of the motor acting, taking the transmission component of the actuator prototype as the target transmission component, and collecting the test state parameters of the target transmission component at different historical moments according to a preset collection frequency, until the motor stops moving or the motor operation accuracy is lower than a preset accuracy threshold, stopping collecting the test state parameters to obtain the test state parameters of the target transmission component at different historical moments; the model of the actuator prototype is the same as the model of the electric actuator; judging whether the test state parameters conform to a preset performance degradation law, and calculating the aging acceleration factor of the target transmission component under the test environment based on the test state parameters conforming to the preset performance degradation law; based on the aging acceleration factor, converting the test state parameters conforming to the preset performance degradation law to obtain the use state parameters of the target transmission component at different historical moments; and using the use state parameters of the target transmission component at different historical moments as the transmission component sample data.

[0014] According to one aspect of an embodiment of the present invention, there is provided an electric actuator performance prediction device, the device comprising: a parameter acquisition module, for acquiring current state parameters of the electric actuator; the electric actuator comprising a transmission component and a circuit board; a component performance prediction module, for inputting the current state parameters of the transmission component into a transmission component performance prediction model, to obtain a performance prediction result of the transmission component at a future time, the transmission component performance prediction model being obtained by training a preset transmission component performance prediction model with transmission component sample data; a circuit performance prediction module, for inputting the current state parameters of the circuit board into a circuit board performance prediction model, to obtain a performance prediction result of the circuit board at a future time; the circuit board performance prediction model being obtained by training a preset circuit board performance prediction model with circuit board sample data; and a result determination module, for determining a performance prediction result of the electric actuator at a future time according to the performance prediction result of the transmission component at a future time and the performance prediction result of the circuit board at a future time.

[0015] According to one aspect of an embodiment of the present invention, there is provided an electronic device, comprising: one or more processors; a storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the electronic device implements the electric actuator performance prediction method as described above.

[0016] According to one aspect of an embodiment of the present application, a computer storage medium is provided, including: computer-readable instructions are stored thereon, and when the computer-readable instructions are executed by a processor of a computer, the computer executes the electric actuator performance prediction method described above.

[0017] Beneficial effects of the present invention: The present invention obtains the current state parameters of the electric actuator, inputs the current state parameters of the transmission component into the transmission component performance prediction model, obtains the performance prediction result of the transmission component at the future moment, inputs the current state parameters of the circuit board into the circuit board performance prediction model, obtains the performance prediction result of the circuit board at the future moment, and determines the performance prediction result of the electric actuator at the future moment according to the performance prediction result of the transmission component at the future moment and the performance prediction result of the circuit board at the future moment. In the above process, the performance of the electric actuator at the future moment is predicted in real time according to the current state parameters of the transmission component. When the current state parameters of the electric actuator are normal, the performance degradation of the electric actuator can be predicted in time, and it is convenient for maintenance personnel to timely evaluate the possible faults of the electric actuator, assist maintenance personnel to perform maintenance services in time, and reduce the situation where the electric actuator is shut down due to faults and affects normal process production.

[0018] 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

[0019] The drawings herein are incorporated into the specification and constitute a part of the specification, showing embodiments consistent with the present application, and together with the specification, are used to explain the principles of the present application. Obviously, the drawings described below are only some embodiments of the present application, and for those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0020] Figure 1 is a schematic diagram of an exemplary system architecture shown in an exemplary embodiment of the present application;

[0021] Figure 2 is a flow chart of a method for predicting performance of an electric actuator shown in an exemplary embodiment of the present application;

[0022] Figure 3 is a block diagram of an electric actuator performance prediction device shown in an exemplary embodiment of the present application;

[0023] Figure 4 A schematic diagram of the structure of a computer system suitable for implementing an electronic device of an embodiment of the present application is shown. DETAILED DESCRIPTION

[0024] The following will describe the embodiments of the present invention with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention, not for limiting the scope of protection of the present invention.

[0025] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and thus the drawings only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0026] In the following description, numerous details are discussed to provide a more thorough explanation of the embodiments of the present invention. However, it is obvious to those skilled in the art that the embodiments of the present invention can be implemented without these specific details. In other embodiments, well-known structures and devices are shown in the form of block diagrams rather than in detail to avoid making the embodiments of the present invention difficult to understand.

[0027] Figure 1 It is a schematic diagram of an exemplary system architecture shown in an exemplary embodiment of the present application.

[0028] Reference Figure 1 As shown, the system architecture may include a storage device 101 and a computer device 102. The computer device 102 may be at least one of a desktop graphics processing unit (GPU) computer, a GPU computing cluster, a neural network computer, etc. Relevant technicians may use the computer device 102 to obtain the current state parameters of the electric actuator, input the current state parameters of the transmission component into the transmission component performance prediction model, obtain the performance prediction result of the transmission component at a future time, input the current state parameters of the circuit board into the circuit board performance prediction model, obtain the performance prediction result of the circuit board at a future time, and determine the performance prediction result of the electric actuator at a future time according to the performance prediction result of the transmission component at a future time and the performance prediction result of the circuit board at a future time. The storage device 101 is used to store the current state parameters of the electric actuator. In this embodiment, the storage device 101 uses a random access memory or other device to store the current state parameters of the electric actuator and provide them to the computer device 102 for processing.

[0029] Illustratively, after obtaining the current state parameters of the electric actuator in the storage device 101, the computer device 102 inputs the current state parameters of the transmission component into the transmission component performance prediction model to obtain the performance prediction result of the transmission component at the future moment, and inputs the current state parameters of the circuit board into the circuit board performance prediction model to obtain the performance prediction result of the circuit board at the future moment, and determines the performance prediction result of the electric actuator at the future moment based on the performance prediction result of the transmission component at the future moment and the performance prediction result of the circuit board at the future moment. In the above process, the performance of the electric actuator at the future moment is predicted in real time based on the current state parameters of the transmission component. When the current state parameters of the electric actuator are normal, the performance degradation of the electric actuator can be predicted in time, and it is convenient for maintenance personnel to timely evaluate possible faults of the electric actuator, assist maintenance personnel to perform maintenance services in time, and reduce the situation where the electric actuator is shut down due to faults and affects normal process production.

[0030] It should be noted that the electric actuator performance prediction method provided in the embodiment of the present application is generally executed by the computer device 102 , and accordingly, the electric actuator performance prediction device is generally disposed in the computer device 102 .

[0031] The implementation details of the technical solution of the embodiment of the present application are described in detail below:

[0032] Figure 2 is a flowchart of an electric actuator performance prediction method shown in an exemplary embodiment of the present application. The electric actuator performance prediction method can be executed by a computing and processing device. The computing and processing device can be Figure 1 The computer device 102 shown in FIG. Figure 2 As shown, the electric actuator performance prediction method at least includes steps S210 to S240, which are described in detail as follows:

[0033] In step S210, the current state parameters of the electric actuator are obtained. In one embodiment of the present application, the electric actuator includes an electric multi-rotation actuator and an electric single-rotation actuator, etc., and the electric actuator includes a transmission component and a circuit board. The current state parameters of the transmission component include the current ambient temperature, the current ambient humidity, the accumulated working time of the motor as of the current time, the accumulated total number of motor movement directions and the accumulated total number of motor movement frequencies as of the current time, the torque size in the motor movement as of the current time, the direction and frequency of the vibration borne by the motor as of the current time, etc., and the current state parameters of the circuit board include the current ambient temperature, the current ambient humidity, the accumulated power-on time of the electric actuator as of the current time, the accumulated working time of the motor as of the current time, the direction and frequency of the vibration borne by the motor as of the current time, etc. The electric actuator is equipped with a temperature sensor and a humidity sensor on the outside, a vibration sensor and monitoring components for various status signals such as the torque size in the motor movement, the accumulated working time of the motor, the motor movement direction and the motor movement frequency. The embedded software inside the electric actuator collects and processes the status signals collected by the sensors and the status signals monitored by the monitoring components to obtain the current status parameters of the electric actuator, and sends the current status parameters of the electric actuator to the electric actuator performance prediction analysis software through the Ethernet digital communication port. The electric actuator performance prediction analysis software determines the performance prediction results of the electric actuator at future times based on the current status parameters of the electric actuator.

[0034] In step S220, the current state parameters of the transmission component are input into the transmission component performance prediction model to obtain the performance prediction result of the transmission component at a future time. In this embodiment, the future moment is a moment reached after a few minutes, hours or days with the current moment as the reference point. The transmission component performance prediction model is obtained by training a preset transmission component performance prediction model with transmission component sample data. The transmission component sample data includes the usage status parameters of the target transmission component at different historical moments. The model of the target transmission component is the same as the model of the transmission component. The preset transmission component performance prediction model is trained with the transmission component sample data to obtain the transmission component performance prediction model. The process includes: inputting the usage status parameters of the target transmission component at different historical moments into a preset feature extraction model to obtain the frequency domain feature signals and time domain feature signals of the target transmission component at different historical moments; training the preset time series prediction model with the frequency domain feature signals of the target transmission component at different historical moments to obtain the trained time series prediction model; inputting the frequency domain feature signals of the target transmission component at different historical moments into the trained time series prediction model to obtain the linear prediction results of the target transmission component at different historical moments; and determining the target transmission component at different historical moments based on the usage status parameters of the target transmission component at different historical moments and the linear prediction results of the target transmission component at different historical moments. The linear residual of the target transmission component at different historical moments is used to train the preset bidirectional long short-term memory neural network model to obtain the trained bidirectional long short-term memory neural network model; the frequency domain characteristic signals of the target transmission component at different historical moments are nonlinearly predicted by the trained bidirectional long short-term memory neural network model to obtain the nonlinear prediction results of the target transmission component at different historical moments; the linear prediction results of the target transmission component at different historical moments and the nonlinear prediction results of the target transmission component at different historical moments are fused by the preset fusion model to obtain the fusion prediction results of the target transmission component at different historical moments, and the fusion prediction results of the target transmission component are represented by the target frequency domain signal; the target frequency domain signal of the target transmission component at different historical moments and the time domain characteristic signal of the target transmission component at different historical moments are transformed by the preset transformation model to obtain the performance prediction results of the target transmission component at different historical moments; based on the performance prediction results of the target transmission component at different historical moments, the preset feature extraction model, the trained time series prediction model, the trained bidirectional long short-term memory neural network model, the preset fusion model and the preset transformation model are combined to obtain the transmission component performance prediction model. The type of the use state parameter of the target transmission component is the same as the type of the current state parameter of the transmission component.

[0035] In step S230, the current state parameters of the circuit board are input into the circuit board performance prediction model to obtain the performance prediction result of the circuit board at a future moment. In this embodiment, the circuit board performance prediction model is obtained by training a preset circuit board performance prediction model with circuit board sample data. The circuit board sample data includes the usage state parameters of the target circuit board at different historical moments. The model of the target circuit board is the same as the model of the circuit board. The preset circuit board performance prediction model is trained with the circuit board sample data to obtain the circuit board performance prediction model. The process includes: inputting the usage state parameters of the target circuit board at different historical moments into a preset feature extraction model to obtain the frequency domain feature signal and time domain feature signal of the target circuit board at different historical moments; training the preset time series prediction model with the frequency domain feature signal of the target circuit board at different historical moments to obtain the first trained time series prediction model; inputting the frequency domain feature signal of the target circuit board at different historical moments into the first trained time series prediction model to obtain the linear prediction result of the target circuit board at different historical moments; and determining the linear residual of the target circuit board at different historical moments based on the usage state parameters of the target circuit board at different historical moments and the linear prediction result of the target circuit board at different historical moments; and determining the linear residual of the target circuit board at different historical moments by the linear residual of the target circuit board at different historical moments. A preset bidirectional long short-term memory neural network model is trained to obtain a first trained bidirectional long short-term memory neural network model; the first trained bidirectional long short-term memory neural network model is used to perform nonlinear prediction on the frequency domain characteristic signals of the target circuit board at different historical moments to obtain the nonlinear prediction results of the target circuit board at different historical moments; the linear prediction results of the target circuit board at different historical moments and the nonlinear prediction results of the target circuit board at different historical moments are fused by a preset fusion model to obtain the fusion prediction results of the target circuit board at different historical moments, and the fusion prediction results of the target circuit board are represented by the first target frequency domain signal; the first target frequency domain signal of the target circuit board at different historical moments and the time domain characteristic signal of the target circuit board at different historical moments are transformed by a preset transformation model to obtain the performance prediction results of the target circuit board at different historical moments; based on the performance prediction results of the target circuit board at different historical moments, the preset feature extraction model, the first trained time series prediction model, the first trained bidirectional long short-term memory neural network model, the preset fusion model and the preset transformation model are combined to obtain the circuit board performance prediction model. The type of the use status parameter of the target circuit board is the same as the type of the current status parameter of the circuit board.

[0036] In step S240, the performance prediction result of the electric actuator at the future moment is determined according to the performance prediction result of the transmission component at the future moment and the performance prediction result of the circuit board at the future moment. In this embodiment, the performance prediction result of the electric actuator at the future moment includes the stroke output accuracy or action execution accuracy of the electric actuator at the future moment. According to the performance prediction result of the transmission component at the future moment and the performance prediction result of the circuit board at the future moment, the process of determining the performance prediction result of the electric actuator at the future moment includes: multiplying the performance prediction result of the transmission component at the future moment with the preset transmission component result weight to obtain a third product; multiplying the performance prediction result of the circuit board at the future moment with the preset circuit board result weight to obtain a fourth product; taking the sum of the third product and the fourth product as the performance prediction result of the electric actuator at the future moment, and performing real-time prediction of the performance of the electric actuator at the future moment according to the current state parameters of the transmission component. When the current state parameters of the electric actuator are normal, the performance degradation of the electric actuator can be predicted in time, and it is convenient for maintenance personnel to timely evaluate the possible faults of the electric actuator, assist maintenance personnel to perform maintenance services in time, and reduce the situation where the electric actuator is shut down due to faults and affects normal process production.

[0037] In one embodiment of the present application, the process of training a preset transmission component performance prediction model by transmission component sample data to obtain the transmission component performance prediction model includes:

[0038] The transmission component sample data is input into a preset feature extraction model to obtain the frequency domain characteristic signal and the time domain characteristic signal of the target transmission component at different historical moments. In this embodiment, the transmission component sample data includes the use status parameters of the target transmission component at different historical moments. The use status parameters of the target transmission component at different historical moments can be the use status parameters of the target transmission component in the entire life cycle, and can also be obtained by converting the test status parameters of the target transmission component at different historical moments in the entire life cycle. The use status parameters of the target transmission component in the entire life cycle are obtained by collecting and processing the status signals collected by the sensor and the status signals monitored by the monitoring component through the embedded software inside the electric actuator. The model of the target transmission component is the same as the model of the transmission component. The process of inputting the transmission component sample data into the preset feature extraction model to obtain the frequency domain feature signal and time domain feature signal of the target transmission component at different historical moments includes: calculating the root mean square of the transmission component sample data, and taking the root mean square of the transmission component sample data as the time domain feature signal of the target transmission component at different historical moments; performing empirical wavelet transform on the time domain feature signal of the target transmission component at different historical moments to obtain the low frequency sub-layer signal of the time domain feature signal of the target transmission component at different historical moments, the low frequency being less than the preset frequency threshold; performing normalization calculation on the low frequency sub-layer signal of the time domain feature signal of the target transmission component at different historical moments to obtain the frequency domain feature signal of the target transmission component at different historical moments. The preset feature extraction model can be divided into a root mean square calculation module, an empirical wavelet transform module and a normalization calculation module, the root mean square calculation module is used to calculate the root mean square of the transmission component sample data; the empirical wavelet transform module is used to perform empirical wavelet transform on the time domain feature signal of the target transmission component at different historical moments; the normalization calculation module is used to perform normalization calculation on the low frequency sub-layer signal of the time domain feature signal of the target transmission component at different historical moments. The process of inputting the usage status parameters of the target circuit board at different historical moments into the preset feature extraction model to obtain the frequency domain characteristic signal and time domain characteristic signal of the target circuit board at different historical moments is the same as the process of inputting the transmission component sample data into the preset feature extraction model to obtain the frequency domain characteristic signal and time domain characteristic signal of the target transmission component at different historical moments.

[0039] The preset time series prediction model is trained by the frequency domain characteristic signals of the target transmission component at different historical moments to obtain the trained time series prediction model. In this embodiment, the preset time series prediction model can select the differential autoregressive integrated moving average (ARIMA) model, or other models can be selected, which are not specifically limited here. A time series refers to a series of values ​​of the same statistical indicator arranged in the order of their occurrence. The main purpose of time series prediction is to predict the frequency domain characteristic signals of the target transmission component at different historical moments based on the frequency domain characteristic signals of the target transmission component at different historical moments. The process of training the preset time series prediction model by the frequency domain characteristic signals of the target circuit board at different historical moments to obtain the first trained time series prediction model is the same as the process of training the preset time series prediction model by the frequency domain characteristic signals of the target transmission component at different historical moments to obtain the trained time series prediction model.

[0040] Input the frequency domain characteristic signals of the target transmission component at different historical moments into the trained time series prediction model to obtain the linear prediction results of the target transmission component at different historical moments; and determine the linear residuals of the target transmission component at different historical moments based on the use state parameters of the target transmission component at different historical moments and the linear prediction results of the target transmission component at different historical moments. In this embodiment, the linear prediction results of the target transmission component at different historical moments include the state parameters of the target transmission component at different historical moments obtained by prediction. Based on the use state parameters of the target transmission component at different historical moments and the linear prediction results of the target transmission component at different historical moments, the process of determining the linear residuals of the target transmission component at different historical moments includes: calculating the difference between the use state parameters and the linear prediction results belonging to the same historical moment, until the difference between the use state parameters and the linear prediction results at all historical moments is calculated, and the linear residuals of the target transmission component at different historical moments are obtained. Input the frequency domain characteristic signals of the target circuit board at different historical moments into the first trained time series prediction model to obtain the linear prediction results of the target circuit board at different historical moments is the same as the process of inputting the frequency domain characteristic signals of the target transmission component at different historical moments into the trained time series prediction model to obtain the linear prediction results of the target transmission component at different historical moments. The process of determining the linear residuals of the target circuit board at different historical moments based on the usage status parameters of the target circuit board at different historical moments and the linear prediction results of the target circuit board at different historical moments is the same as the process of determining the linear residuals of the target transmission component at different historical moments based on the usage status parameters of the target transmission component at different historical moments and the linear prediction results of the target transmission component at different historical moments.

[0041] The preset bidirectional long short-term memory neural network model is trained by the linear residuals of the target transmission component at different historical moments to obtain the trained bidirectional long short-term memory neural network model. In this embodiment, before the preset bidirectional long short-term memory neural network model is trained by the linear residuals of the target transmission component at different historical moments, the linear residuals of the target transmission component at different historical moments need to be subjected to time window processing. The process of training the preset bidirectional long short-term memory neural network model by the linear residuals of the target transmission component at different historical moments is the process of adjusting the number of neurons, the number of neural network layers and the learning rate in the preset bidirectional long short-term memory neural network model. The process of training the preset bidirectional long short-term memory neural network model by the linear residuals of the target circuit board at different historical moments to obtain the first trained bidirectional long short-term memory neural network model is the same as the process of training the preset bidirectional long short-term memory neural network model by the linear residuals of the target transmission component at different historical moments to obtain the trained bidirectional long short-term memory neural network model.

[0042] The frequency domain characteristic signals of the target transmission component at different historical moments are nonlinearly predicted by the trained bidirectional long short-term memory neural network model to obtain the nonlinear prediction results of the target transmission component at different historical moments. In this embodiment, the preset bidirectional long short-term memory (Bidirectional Long Short-Term Memory, Bi-LSTM) neural network model is suitable for processing time series data with correlation before and after, and at the same time, it makes up for the limitations of processing time series data with a unidirectional propagation structure. The process of performing nonlinear prediction on the frequency domain characteristic signals of the target transmission component at different historical moments by the trained bidirectional long short-term memory neural network model to obtain the nonlinear prediction results of the target transmission component at different historical moments is the same as the process of performing nonlinear prediction on the frequency domain characteristic signals of the target circuit board at different historical moments by the first trained bidirectional long short-term memory neural network model to obtain the nonlinear prediction results of the target circuit board at different historical moments.

[0043] The linear prediction results of the target transmission component at different historical moments and the nonlinear prediction results of the target transmission component at different historical moments are fused using a preset fusion model to obtain the fused prediction results of the target transmission component at different historical moments, and the fused prediction results of the target transmission component are represented by a target frequency domain signal. In this embodiment, the process of fusion of the linear prediction results of the target transmission component at different historical moments and the nonlinear prediction results of the target transmission component at different historical moments using a preset fusion model to obtain the fused prediction results of the target transmission component at different historical moments includes: selecting a moment as a target moment from different historical moments, multiplying the linear prediction result of the target transmission component at the target moment by a preset linear result weight to obtain a first product; multiplying the nonlinear prediction result of the target transmission component at the target moment by a preset nonlinear result weight to obtain a second product; taking the sum of the first product and the second product as the fused prediction result of the target transmission component at the target moment, reselecting a moment from different historical moments, and calculating the fused prediction result of the target transmission component at the corresponding moment, until all historical moments are selected and the calculation is terminated to obtain the fused prediction results of the target transmission component at different historical moments. The process of using a preset fusion model to fuse the linear prediction results of the target circuit board at different historical moments and the nonlinear prediction results of the target circuit board at different historical moments to obtain the fused prediction results of the target circuit board at different historical moments is the same as the process of using a preset fusion model to fuse the linear prediction results of the target transmission component at different historical moments and the nonlinear prediction results of the target transmission component at different historical moments to obtain the fused prediction results of the target transmission component at different historical moments.

[0044] The target frequency domain signals of the target transmission component at different historical moments and the time domain characteristic signals of the target transmission component at different historical moments are transformed by using a preset transformation model to obtain performance prediction results of the target transmission component at different historical moments. In this embodiment, the target frequency domain signals of the target transmission component at different historical moments and the time domain characteristic signals of the target transmission component at different historical moments are transformed to obtain performance prediction results of the target transmission component at different historical moments. The process includes: combining the target frequency domain signals and the time domain characteristic signals belonging to the same historical moment in the target transmission component until the combination of the target frequency domain signals and the time domain characteristic signals of all historical moments in the target transmission component is completed and the combination is terminated to obtain combined characteristic signals of the target transmission component at different historical moments; performing an empirical wavelet inverse transform on the combined characteristic signals of the target transmission component at different historical moments to obtain performance prediction results of the target transmission component at different historical moments. The process of using a preset transformation model to transform the first target frequency domain signal of the target circuit board at different historical moments and the time domain characteristic signal of the target circuit board at different historical moments to obtain the performance prediction results of the target circuit board at different historical moments is the same as the process of using a preset transformation model to transform the target frequency domain signal of the target transmission component at different historical moments and the time domain characteristic signal of the target transmission component at different historical moments to obtain the performance prediction results of the target transmission component at different historical moments.

[0045] Based on the performance prediction results of the target circuit board at different historical moments, the preset feature extraction model, the trained time series prediction model, the trained bidirectional long short-term memory neural network model, the preset fusion model and the preset transformation model are combined to obtain a transmission component performance prediction model. In this embodiment, the target frequency domain signal of the target transmission component at different historical moments and the time domain feature signal of the target transmission component at different historical moments are transformed by using an empirical wavelet inverse transformation, etc. The empirical wavelet inverse transformation process is the inverse process of the empirical wavelet transformation.

[0046] In one embodiment of the present application, the process of inputting the transmission component sample data into a preset feature extraction model to obtain the frequency domain feature signal and the time domain feature signal of the target transmission component at different historical moments includes:

[0047] The root mean square of the transmission component sample data is calculated, and the root mean square of the transmission component sample data is used as the time domain characteristic signal of the target transmission component at different historical moments. In this embodiment, the time domain characteristic signal of the target transmission component at different historical moments can also be characterized by the mean, variance, skewness or kurtosis of the transmission component sample data.

[0048] The time domain characteristic signals of the target transmission component at different historical moments are subjected to empirical wavelet transform to obtain low-frequency sub-layer signals of the time domain characteristic signals of the target transmission component at different historical moments. In this embodiment, the low frequency is less than a preset frequency threshold, and the preset frequency threshold can be set according to actual conditions and is not limited here.

[0049] The low-frequency sub-layer signals of the time-domain characteristic signals of the target transmission component at different historical moments are normalized and calculated to obtain the frequency-domain characteristic signals of the target transmission component at different historical moments. In this embodiment, the low-frequency sub-layer signals of the time-domain characteristic signals of the target transmission component at different historical moments are normalized and calculated, which has the effect of simplifying calculations, improving data comparability, and improving the convergence speed and prediction accuracy of the preset feature extraction model.

[0050] In one embodiment of the present application, the process of fusing the linear prediction results of the target transmission component at different historical moments and the nonlinear prediction results of the target transmission component at different historical moments to obtain the fused prediction results of the target transmission component at different historical moments includes:

[0051] A moment is selected from different historical moments as a target moment, and a linear prediction result of a target transmission component at the target moment is multiplied by a preset linear result weight to obtain a first product; a nonlinear prediction result of a target transmission component at the target moment is multiplied by a preset nonlinear result weight to obtain a second product; and a sum of the first product and the second product is used as a fusion prediction result of the target transmission component at the target moment. In this embodiment, the preset linear result weight is set according to actual conditions, and the preset nonlinear result weight is set according to actual conditions, and no specific limitation is made here.

[0052] A moment is selected and reselected from different historical moments, and the fusion prediction result of the target transmission component at the corresponding moment is calculated, until all historical moments are selected and the calculation is terminated, so as to obtain the fusion prediction result of the target transmission component at different historical moments. In this embodiment, the process of calculating the fusion prediction result of the target transmission component at the corresponding moment is calculated according to the process of calculating the fusion prediction result of the target moment, and no repeated limitation is made here.

[0053] In one embodiment of the present application, the process of transforming the target frequency domain signal of the target transmission component at different historical moments and the time domain characteristic signal of the target transmission component at different historical moments to obtain the performance prediction results of the target transmission component at different historical moments includes:

[0054] The target frequency domain signal and the time domain characteristic signal belonging to the same historical moment in the target transmission component are combined until the target frequency domain signal and the time domain characteristic signal of all historical moments in the target transmission component are combined to obtain the combined characteristic signal of different historical moments in the target transmission component. In this embodiment, the fusion prediction results and the time domain characteristic signals of all historical moments in the target transmission component are combined to avoid the loss of the characteristic signals using the state parameters in the target transmission component.

[0055] An empirical wavelet inverse transform is performed on the combined characteristic signals of the target transmission component at different historical moments to obtain the performance prediction results of the target transmission component at different historical moments. In this embodiment, the empirical wavelet transform can remove the noise components in the signal, and after the combined characteristic signals of the target transmission component at different historical moments are subjected to the empirical wavelet inverse transform, it is beneficial to the noise suppression and signal enhancement in the combined characteristic signals of the target transmission component at different historical moments.

[0056] In one embodiment of the present application, the process of determining the performance prediction result of the electric actuator at a future time according to the performance prediction result of the transmission component at a future time and the performance prediction result of the circuit board at a future time includes:

[0057] The performance prediction result of the transmission component at the future moment is multiplied by the preset transmission component result weight to obtain a third product. In this embodiment, the preset transmission component result weight is set according to actual conditions and is not specifically limited here.

[0058] The performance prediction result of the circuit board at the future moment is multiplied by the preset circuit board result weight to obtain a fourth product. In this embodiment, the preset circuit board result weight can be set according to actual conditions and is not specifically limited here.

[0059] The sum of the third product and the fourth product is used as the performance prediction result of the electric actuator at the future time. In this embodiment, the performance prediction result of the electric actuator at the future time is embodied in the form of a numerical value, that is, the higher the numerical value of the performance prediction result at the future time, the better the performance of the electric actuator, and the lower the numerical value of the performance prediction result at the future time, the worse the performance of the electric actuator.

[0060] In one embodiment of the present application, before training a preset transmission component performance prediction model through transmission component sample data to obtain the transmission component performance prediction model, the electric actuator performance prediction method further includes:

[0061] The motor of the actuator prototype is controlled to act under a preset test environment, and, in the process of the motor of the actuator prototype acting, the transmission component of the actuator prototype is used as the target transmission component, and the test state parameters of the target transmission component at different historical moments are collected according to the preset collection frequency, until the motor stops moving or the motor operation accuracy is lower than the preset accuracy threshold, then the collection of the test state parameters of the target transmission component is stopped to obtain the test state parameters of the target transmission component at different historical moments. In this embodiment, the preset accuracy threshold can be set according to actual conditions, the model of the actuator prototype is the same as the model of the electric actuator, the preset test environment can be a test environment in a high temperature test box, or it can be other test environments, the preset test environment includes a preset ambient temperature, a preset ambient humidity and a preset installation position, etc. The preset ambient temperature can be a fixed ambient temperature or an ambient temperature that changes according to a preset change rule, the preset ambient humidity can be a fixed ambient humidity or an ambient humidity that changes according to a preset change rule, and the preset installation position can be a fluid pipeline or other positions. During the operation of the motor of the actuator prototype, the fluid pipeline applies different vibration intensities to the actuator prototype, and the motor of the actuator prototype moves continuously in different directions and at different positions.

[0062] Determine whether the test state parameters of the target transmission component at different historical moments conform to the preset performance decay law, and calculate the aging acceleration factor of the target transmission component under the test environment based on the test state parameters of the target transmission component that conforms to the preset performance decay law. In this embodiment, the process of determining whether the test state parameters of the target transmission component at different historical moments conform to the preset performance decay law includes: inputting the test state parameters of the target transmission component at different historical moments into the Wiener process model to obtain the test state parameters of the target transmission component that conforms to the preset performance decay law, and the preset performance decay law is the accelerated failure mechanism of the electric actuator. Based on the test state parameters of the target transmission component that conforms to the preset performance decay law, the process of calculating the aging acceleration factor of the target transmission component under the test environment includes: inputting the test state parameters of the target transmission component that conforms to the preset performance decay law into the Arrhenius equation to calculate the aging acceleration factor of the target transmission component under the test environment.

[0063] Based on the aging acceleration factor of the target transmission component in the test environment, the test state parameters of the target transmission component that conforms to the preset performance degradation law are converted to obtain the usage state parameters of the target transmission component at different historical moments. In this embodiment, based on the aging acceleration factor of the target transmission component in the test environment, the test state parameters of the target transmission component that conforms to the preset performance degradation law are converted to obtain the usage state parameters of the target transmission component at different historical moments. The process can be implemented by referring to the process of converting the test state parameters into the usage state parameters through the aging acceleration factor in the related art. The usage state parameters are the state parameters of the electric actuator during use.

[0064] The usage state parameters of the target transmission component at different historical moments are used as the transmission component sample data. In this embodiment, after the transmission component sample data is obtained, the test state parameters of other target transmission components can be converted into usage state parameters and added to the original transmission component sample data, thereby enriching and updating the transmission component sample data, and then continuously improving the prediction accuracy of the transmission component performance prediction model.

[0065] In this embodiment, before the preset circuit board performance prediction model is trained by circuit board sample data to obtain the circuit board performance prediction model, the electric actuator performance prediction method also includes: controlling the motor of the actuator prototype to operate under a preset test environment, and, during the operation of the motor of the actuator prototype, performing a target circuit board operation on the circuit board of the actuator prototype, and collecting the test state parameters of the target circuit board at different historical moments according to a preset acquisition frequency, until the motor stops moving or the motor operation accuracy is lower than a preset accuracy threshold, stopping the collection of the test state parameters of the target circuit board to obtain the test state parameters of the target circuit board at different historical moments; judging whether the test state parameters of the target circuit board at different historical moments conform to a preset performance degradation law, and calculating the aging acceleration factor of the target circuit board under the test environment based on the test state parameters of the target circuit board that conforms to the preset performance degradation law; based on the aging acceleration factor of the target circuit board under the test environment, converting the test state parameters of the target circuit board that conforms to the preset performance degradation law to obtain the usage state parameters of the target circuit board at different historical moments; and using the usage state parameters of the target circuit board at different historical moments as circuit board sample data.

[0066] In this embodiment, the process of judging whether the test state parameters of the target circuit board at different historical moments conform to the preset performance degradation law, and calculating the aging acceleration factor of the target circuit board under the test environment based on the test state parameters of the target circuit board that conform to the preset performance degradation law is the same as the process of judging whether the test state parameters of the target transmission component at different historical moments conform to the preset performance degradation law, and calculating the aging acceleration factor of the target transmission component under the test environment based on the test state parameters of the target transmission component that conforms to the preset performance degradation law.

[0067] In this embodiment, based on the aging acceleration factor of the target circuit board in the test environment, the test state parameters of the target circuit board that conforms to the preset performance degradation law are converted to obtain the use state parameters of the target circuit board at different historical moments. The process is the same as the process of converting the test state parameters of the target transmission component that conforms to the preset performance degradation law based on the aging acceleration factor of the target transmission component in the test environment to obtain the use state parameters of the target transmission component at different historical moments.

[0068] In one embodiment of the present application, first, historical state parameters are collected by the intelligent electric actuator, and the historical state parameters include the use state parameters of the target transmission component obtained by directly collecting and the use state parameters obtained by converting the test state parameters of the target transmission component; then, the historical state parameters are stored in the database of the expert system, and the expert system includes a transmission component performance prediction model, a circuit board performance prediction model, a database and a computer, wherein the transmission component performance prediction model is obtained by training a preset transmission component performance prediction model through the historical state data of the target transmission component; the circuit board performance prediction model is obtained by training a preset circuit board performance prediction model through the historical state data of the target circuit board; the database is a time series database; the computer is a storage and calculation carrier of software such as the transmission component performance prediction model, the circuit board performance prediction model, and the database; Finally, through the real-time status parameters collected by the intelligent electric actuator, and the performance prediction model of the transmission component and the circuit board performance prediction model in the expert system are called by the performance prediction analysis software, the real-time status parameters of the transmission component are input into the transmission component performance prediction model to obtain the performance prediction result of the transmission component at the future moment; the real-time status parameters of the circuit board are input into the circuit board performance prediction model to obtain the performance prediction result of the circuit board at the future moment; the performance prediction results of the transmission component at the future moment and the performance prediction results of the circuit board at the future moment are weightedly summed to obtain the overall performance prediction result of the intelligent electric actuator at the future moment, and the overall performance prediction result of the intelligent electric actuator at the future moment is presented on the corresponding human-machine interface. The performance prediction analysis software is a host computer monitoring software composed of a series of network communications, function scheduling and human-machine interfaces.

[0069] The following describes an embodiment of the device of the present application, which can be used to implement the electric actuator performance prediction method in the above embodiment of the present application. For details not disclosed in the embodiment of the device of the present application, please refer to the embodiment of the electric actuator performance prediction method in the above embodiment of the present application.

[0070] Figure 3 is a block diagram of an electric actuator performance prediction device shown in an exemplary embodiment of the present application. The device can be applied to Figure 1 The implementation environment shown in the figure is specifically configured in the computer device 102. The device can also be applied to other exemplary implementation environments and specifically configured in other devices. This embodiment does not limit the implementation environment to which the device is applied.

[0071] like Figure 3 As shown, the exemplary electric actuator performance prediction device includes:

[0072] The parameter acquisition module 301 is used to acquire the current state parameters of the electric actuator; the electric actuator includes a transmission component and a circuit board.

[0073] The component performance prediction module 302 is used to input the current state parameters of the transmission component into the transmission component performance prediction model to obtain the performance prediction result of the transmission component at a future moment.

[0074] The circuit performance prediction module 303 is used to input the current state parameters of the circuit board into the circuit board performance prediction model to obtain the performance prediction result of the circuit board at a future time.

[0075] The result determination module 304 is used to determine the performance prediction result of the electric actuator at a future time according to the performance prediction result of the transmission component at a future time and the performance prediction result of the circuit board at a future time.

[0076] In one embodiment of the present application, the electric actuator includes an electric multi-rotation actuator and an electric single-rotation actuator, etc., and the electric actuator includes a transmission component and a circuit board. The current state parameters of the transmission component include the current ambient temperature, the current ambient humidity, the accumulated working time of the motor as of the current time, the accumulated total number of motor movement directions and the accumulated total number of motor movement frequencies as of the current time, the torque magnitude in the motor movement as of the current time, the direction and frequency of vibration suffered in the motor movement as of the current time, etc., and the current state parameters of the circuit board include the current ambient temperature, the current ambient humidity, the accumulated power-on time of the electric actuator as of the current time, the accumulated working time of the motor as of the current time, the direction and frequency of vibration suffered in the motor movement as of the current time, etc. The electric actuator is equipped with a temperature sensor and a humidity sensor on the outside, a vibration sensor and monitoring components for various status signals such as the torque size in the motor movement, the accumulated working time of the motor, the motor movement direction and the motor movement frequency. The embedded software inside the electric actuator collects and processes the status signals collected by the sensors and the status signals monitored by the monitoring components to obtain the current status parameters of the electric actuator, and sends the current status parameters of the electric actuator to the electric actuator performance prediction analysis software through the Ethernet digital communication port. The electric actuator performance prediction analysis software determines the performance prediction results of the electric actuator at future times based on the current status parameters of the electric actuator.

[0077] In one embodiment of the present application, the future moment is a moment reached after a few minutes, hours or days with the current moment as the reference point. The transmission component performance prediction model is obtained by training a preset transmission component performance prediction model with transmission component sample data. The transmission component sample data includes the usage status parameters of the target transmission component at different historical moments. The model of the target transmission component is the same as the model of the transmission component. The preset transmission component performance prediction model is trained with the transmission component sample data. The process of obtaining the transmission component performance prediction model includes: inputting the usage status parameters of the target transmission component at different historical moments into a preset feature extraction model to obtain the frequency domain feature signals and time domain feature signals of the target transmission component at different historical moments; training the preset time series prediction model with the frequency domain feature signals of the target transmission component at different historical moments to obtain the trained time series prediction model; inputting the frequency domain feature signals of the target transmission component at different historical moments into the trained time series prediction model to obtain the linear prediction results of the target transmission component at different historical moments; and determining the target transmission component at different historical moments based on the usage status parameters of the target transmission component at different historical moments and the linear prediction results of the target transmission component at different historical moments. The linear residual at different moments; the preset bidirectional long short-term memory neural network model is trained by the linear residual at different historical moments of the target transmission component to obtain the trained bidirectional long short-term memory neural network model; the frequency domain characteristic signal of the target transmission component at different historical moments is nonlinearly predicted by the trained bidirectional long short-term memory neural network model to obtain the nonlinear prediction results of the target transmission component at different historical moments; the linear prediction results of the target transmission component at different historical moments and the nonlinear prediction results of the target transmission component at different historical moments are fused by the preset fusion model to obtain the fusion prediction results of the target transmission component at different historical moments, and the fusion prediction results of the target transmission component are represented by the target frequency domain signal; the target frequency domain signal of the target transmission component at different historical moments and the time domain characteristic signal of the target transmission component at different historical moments are transformed by the preset transformation model to obtain the performance prediction results of the target transmission component at different historical moments; based on the performance prediction results of the target transmission component at different historical moments, the preset feature extraction model, the trained time series prediction model, the trained bidirectional long short-term memory neural network model, the preset fusion model and the preset transformation model are combined to obtain the transmission component performance prediction model. The type of the use state parameter of the target transmission component is the same as the type of the current state parameter of the transmission component.

[0078] In one embodiment of the present application, a circuit board performance prediction model is obtained by training a preset circuit board performance prediction model with circuit board sample data, the circuit board sample data includes usage status parameters of a target circuit board at different historical moments, the model of the target circuit board is the same as the model of the circuit board, and the preset circuit board performance prediction model is trained with the circuit board sample data, and the process of obtaining the circuit board performance prediction model includes: inputting the usage status parameters of the target circuit board at different historical moments into a preset feature extraction model to obtain frequency domain feature signals and time domain feature signals of the target circuit board at different historical moments; training a preset time series prediction model with the frequency domain feature signals of the target circuit board at different historical moments to obtain a first trained time series prediction model; inputting the frequency domain feature signals of the target circuit board at different historical moments into the first trained time series prediction model to obtain linear prediction results of the target circuit board at different historical moments; and determining the linear residuals of the target circuit board at different historical moments based on the usage status parameters of the target circuit board at different historical moments and the linear prediction results of the target circuit board at different historical moments; and determining the linear residuals of the target circuit board at different historical moments based on the linear residuals of the target circuit board at different historical moments. The preset bidirectional long short-term memory neural network model is trained to obtain the first trained bidirectional long short-term memory neural network model; the frequency domain characteristic signals of the target circuit board at different historical moments are nonlinearly predicted by the first trained bidirectional long short-term memory neural network model to obtain the nonlinear prediction results of the target circuit board at different historical moments; the linear prediction results of the target circuit board at different historical moments and the nonlinear prediction results of the target circuit board at different historical moments are fused by the preset fusion model to obtain the fusion prediction results of the target circuit board at different historical moments, and the fusion prediction results of the target circuit board are represented by the first target frequency domain signal; the first target frequency domain signal of the target circuit board at different historical moments and the time domain characteristic signal of the target circuit board at different historical moments are transformed by the preset transformation model to obtain the performance prediction results of the target circuit board at different historical moments; based on the performance prediction results of the target circuit board at different historical moments, the preset feature extraction model, the first trained time series prediction model, the first trained bidirectional long short-term memory neural network model, the preset fusion model and the preset transformation model are combined to obtain the circuit board performance prediction model. The type of the use status parameter of the target circuit board is the same as the type of the current status parameter of the circuit board.

[0079] In one embodiment of the present application, the performance prediction result of the electric actuator at a future moment includes the stroke output accuracy or action execution accuracy of the electric actuator at a future moment. According to the performance prediction result of the transmission component at a future moment and the performance prediction result of the circuit board at a future moment, the process of determining the performance prediction result of the electric actuator at a future moment includes: multiplying the performance prediction result of the transmission component at a future moment with the preset transmission component result weight to obtain a third product; multiplying the performance prediction result of the circuit board at a future moment with the preset circuit board result weight to obtain a fourth product; taking the sum of the third product and the fourth product as the performance prediction result of the electric actuator at a future moment, and making a real-time prediction of the performance of the electric actuator at a future moment according to the current state parameters of the transmission component. When the current state parameters of the electric actuator are normal, the performance degradation of the electric actuator can be predicted in time, and it is convenient for maintenance personnel to timely evaluate the possible faults of the electric actuator, assist maintenance personnel to perform maintenance services in time, and reduce the situation where the electric actuator is shut down due to faults and affects normal process production.

[0080] It should be noted that the electric actuator performance prediction device provided in the above embodiment and the electric actuator performance prediction method provided in the above embodiment belong to the same concept, wherein the specific manner in which each module and unit performs the operation has been described in detail in the method embodiment, and will not be repeated here. In practical applications, the electric actuator performance prediction device provided in the above embodiment can allocate the above functions to different functional modules as needed, that is, divide the internal structure of the device into different functional modules to complete all or part of the functions described above, and this is not limited here.

[0081] An embodiment of the present application also provides an electronic device, comprising: one or more processors; a storage device for storing one or more programs, and when the one or more programs are executed by one or more processors, the electronic device implements the electric actuator performance prediction method provided in the above-mentioned embodiments.

[0082] Figure 4 The structure diagram of the computer system suitable for implementing the electronic device of the embodiment of the present application is shown. It should be noted that: Figure 4 The computer system 400 of the electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0083] like Figure 4As shown, the computer system 400 includes a central processing unit (CPU) 401, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 402 or the program loaded from the storage part 408 to the random access memory (RAM) 403, such as executing the method in the above embodiment. In the RAM 403, various programs and data required for system operation are also stored. The CPU 401, ROM 402 and RAM 403 are connected to each other through a bus 404. An input / output (I / O) interface 405 is also connected to the bus 404.

[0084] The following components are connected to the I / O interface 405: an input section 406 including a keyboard, a mouse, etc.; an output section 407 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 408 including a hard disk, etc.; and a communication section 409 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 409 performs communication processing via a network such as the Internet. A drive 410 is also connected to the I / O interface 405 as needed. A removable medium 411, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 410 as needed so that a computer program read therefrom is installed into the storage section 408 as needed.

[0085] In particular, according to an embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication section 409, and / or installed from a removable medium 411. When the computer program is executed by a central processing unit (CPU) 401, various functions defined in the system of the present application are executed.

[0086] It should be noted that the computer-readable medium shown in the embodiment of the present application can be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium can be, for example, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media can include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, wherein a computer-readable computer program is carried. This propagated data signal can take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. A computer program contained on a computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0087] The flowchart and block diagram in the accompanying drawings illustrate the possible architecture, functions and operations of the system, method and computer program product according to various embodiments of the present application. Wherein, each box in the flowchart or block diagram can represent a module, a program segment, or a part of the code, and the above-mentioned module, program segment, or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0088] The units involved in the embodiments described in this application may be implemented by software or hardware, and the units described may also be set in a processor. The names of these units do not, in some cases, constitute limitations on the units themselves.

[0089] Another aspect of the present application provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor of a computer, the computer executes the electric actuator performance prediction method as described above. The computer-readable storage medium may be included in the electronic device described in the above embodiment, or may exist independently without being assembled into the electronic device.

[0090] Another aspect of the present application also provides a computer program product or a computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the electric actuator performance prediction method provided in each of the above embodiments.

[0091] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the technology may modify or change the above embodiments without violating the spirit and scope of the present invention. Therefore, all equivalent modifications or changes made by a person of ordinary skill in the art without departing from the spirit and technical ideas disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. A method for predicting performance of an electric actuator, characterized in that: The method comprises: Acquiring current state parameters of an electric actuator; the electric actuator includes a transmission component and a circuit board; Inputting the current state parameters of the transmission component into a transmission component performance prediction model to obtain a performance prediction result of the transmission component at a future time, wherein the transmission component performance prediction model is obtained by training a preset transmission component performance prediction model with transmission component sample data; Inputting the current state parameters of the circuit board into a circuit board performance prediction model to obtain a performance prediction result of the circuit board at a future time; the circuit board performance prediction model is obtained by training a preset circuit board performance prediction model with circuit board sample data; The performance prediction result of the electric actuator at the future moment is determined according to the performance prediction result of the transmission component at the future moment and the performance prediction result of the circuit board at the future moment.

2. The electric actuator performance prediction method according to claim 1, characterized in that: The process of training a preset transmission component performance prediction model by transmission component sample data to obtain the transmission component performance prediction model includes: Input the transmission component sample data into a preset feature extraction model to obtain frequency domain feature signals and time domain feature signals of the target transmission component at different historical moments; the transmission component sample data includes the usage status parameters of the target transmission component at different historical moments; the model of the target transmission component is the same as the model of the transmission component; The preset time series prediction model is trained by using the frequency domain feature signal to obtain a trained time series prediction model; Inputting the frequency domain characteristic signal into the trained time series prediction model to obtain linear prediction results of the target transmission component at different historical moments; and determining the linear residual of the target transmission component at different historical moments based on the usage state parameter and the linear prediction result; The preset bidirectional long short-term memory neural network model is trained by using the linear residual to obtain a trained bidirectional long short-term memory neural network model; Performing nonlinear prediction on the frequency domain characteristic signal by using the trained bidirectional long short-term memory neural network model to obtain nonlinear prediction results of the target transmission component at different historical moments; The linear prediction result and the nonlinear prediction result are fused by using a preset fusion model to obtain fused prediction results of the target transmission component at different historical moments, and the fused prediction results are represented by a target frequency domain signal; The target frequency domain signal and the time domain characteristic signal are transformed by using a preset transformation model to obtain performance prediction results of the target transmission component at different historical moments; Based on the performance prediction results of the target transmission component at different historical moments, the preset feature extraction model, the trained time series prediction model, the trained bidirectional long short-term memory neural network model, the preset fusion model and the preset transformation model are combined to obtain the transmission component performance prediction model.

3. The electric actuator performance prediction method according to claim 2, characterized in that: The process of inputting the transmission component sample data into the preset feature extraction model to obtain the frequency domain feature signal and the time domain feature signal of the target transmission component at different historical moments includes: Calculating the root mean square of the transmission component sample data, and using the root mean square of the transmission component sample data as the time domain characteristic signal of the target transmission component at different historical moments; Performing an empirical wavelet transform on the time domain characteristic signal to obtain a low-frequency sub-layer signal of the time domain characteristic signal, wherein the low frequency is less than a preset frequency threshold; The low-frequency sub-layer signal of the time-domain characteristic signal is normalized and calculated to obtain the frequency-domain characteristic signal of the target transmission component at different historical moments.

4. The electric actuator performance prediction method according to claim 2, characterized in that: The process of fusing the linear prediction result and the nonlinear prediction result to obtain the fused prediction result of the target transmission component at different historical moments includes: Select a moment from different historical moments as a target moment, multiply the linear prediction result of the target moment by a preset linear result weight to obtain a first product; multiply the nonlinear prediction result of the target moment by a preset nonlinear result weight to obtain a second product; and use the sum of the first product and the second product as the fusion prediction result of the target transmission component at the target moment; A moment is reselected from different historical moments, and the fusion prediction result of the target transmission component at the corresponding moment is calculated, until all historical moments are selected and the calculation is ended to obtain the fusion prediction results of the target transmission component at different historical moments.

5. The electric actuator performance prediction method according to claim 2, characterized in that: The process of transforming the target frequency domain signal and the time domain characteristic signal to obtain the performance prediction results of the target transmission component at different historical moments includes: The target frequency domain signal and the time domain characteristic signal belonging to the same historical moment are combined until the combination of the target frequency domain signal and the time domain characteristic signal of all historical moments is completed, thereby obtaining the combined characteristic signal of the target transmission component at different historical moments; An empirical wavelet inverse transform is performed on the combined characteristic signal to obtain performance prediction results of the target transmission component at different historical moments.

6. The electric actuator performance prediction method according to any one of claims 1 to 5, characterized in that: The process of determining the performance prediction result of the electric actuator at a future time according to the performance prediction result of the transmission component at a future time and the performance prediction result of the circuit board at a future time includes: Multiplying the performance prediction result of the transmission component at a future time by a preset transmission component result weight to obtain a third product; Multiplying the performance prediction result of the circuit board at a future time by a preset circuit board result weight to obtain a fourth product; The sum of the third product and the fourth product is used as the performance prediction result of the electric actuator at a future moment.

7. The electric actuator performance prediction method according to any one of claims 2 to 5, characterized in that: Before training the preset transmission component performance prediction model by using the transmission component sample data to obtain the transmission component performance prediction model, the method further includes: The motor of the actuator prototype is controlled to move under a preset test environment, and, in the process of the motor moving, the transmission component of the actuator prototype is used as the target transmission component, and the test state parameters of the target transmission component at different historical moments are collected according to a preset collection frequency, until the motor stops moving or the motor operation accuracy is lower than a preset accuracy threshold, then the collection of the test state parameters is stopped, and the test state parameters of the target transmission component at different historical moments are obtained; the model of the actuator prototype is the same as the model of the electric actuator; Determining whether the test state parameters conform to a preset performance degradation law, and calculating an aging acceleration factor of the target transmission component under the test environment based on the test state parameters conforming to the preset performance degradation law; Based on the aging acceleration factor, the test state parameters that conform to the preset performance degradation law are converted to obtain the use state parameters of the target transmission component at different historical moments; The usage status parameters of the target transmission component at different historical moments are used as the transmission component sample data.

8. An electric actuator performance prediction device, characterized in that: The device comprises: A parameter acquisition module, used to acquire current state parameters of the electric actuator; the electric actuator includes a transmission component and a circuit board; A component performance prediction module, used for inputting the current state parameters of the transmission component into a transmission component performance prediction model to obtain a performance prediction result of the transmission component at a future time, wherein the transmission component performance prediction model is obtained by training a preset transmission component performance prediction model with transmission component sample data; A circuit performance prediction module, used to input the current state parameters of the circuit board into a circuit board performance prediction model to obtain a performance prediction result of the circuit board at a future time; the circuit board performance prediction model is obtained by training a preset circuit board performance prediction model with circuit board sample data; The result determination module is used to determine the performance prediction result of the electric actuator at a future moment according to the performance prediction result of the transmission component at a future moment and the performance prediction result of the circuit board at a future moment.

9. An electronic device, characterized in that: include: one or more processors; A storage device for storing one or more programs, which, when executed by the one or more processors, enables the electronic device to implement the electric actuator performance prediction method as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that: include: Computer-readable instructions are stored thereon, and when the computer-readable instructions are executed by a processor of a computer, the computer is caused to execute the electric actuator performance prediction method according to any one of claims 1 to 7.

Images