Energy Consumption Optimization Method and System for Electric Actuators Based on Big Data Analysis

Abstract

The present invention discloses an energy consumption optimization method and system for electric actuators based on big data analysis, specifically relating to the technical field of energy consumption optimization; by collecting multi-source operation data, using a sliding window to extract time series features to generate working condition modal feature vectors, and combining with an energy efficiency influence factor weight matrix to construct an energy efficiency influence factor set. Using a dynamic density clustering algorithm to perform working condition state clustering on the energy efficiency influence factor set to generate an operation mode classification tree; conducting energy consumption traceability analysis on the non-steady operation mode to generate an energy efficiency degradation feature map. Subsequently, input the energy efficiency degradation feature map into a pre-trained energy consumption optimization decision model, output a dynamic regulation strategy set, and after the strategy takes effect, by comparing the deviation degree between the energy efficiency transition trajectory and the theoretical optimal curve, iteratively update the model weights, and output the final optimized control parameter package to achieve intelligent optimization and energy-saving control of the operation mode of the electric actuator.

Description

Technical Field

[0001] The present invention relates to the technical field of energy consumption optimization, and more specifically, to an energy consumption optimization method and system for electric actuators based on big data analysis. Background Art

[0002] Traditional energy consumption management methods for electric actuators usually rely on static parameter settings and simple control strategies, making it difficult to cope with energy consumption fluctuations under variable working conditions and lacking full utilization of real-time operation data. Existing technologies have deficiencies in data fusion, time series feature extraction, and working condition state analysis, and are unable to accurately describe the complex relationship between the actuator operation mode and energy efficiency decay, resulting in a lag in the response of the regulation strategy and limited energy consumption optimization effect.

[0003] To solve the above problems, a technical solution is provided now. Summary of the Invention

[0004] To overcome the above defects of the prior art, embodiments of the present invention provide an energy consumption optimization method and system for electric actuators based on big data analysis to solve the problems raised in the above background art.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] An energy consumption optimization method for electric actuators based on big data analysis, comprising the following steps:

[0007] Collect multi-source operation data streams of the electric actuator, and perform time series feature extraction on the multi-source operation data streams based on a sliding window to generate an actuator working condition mode feature vector;

[0008] According to the actuator working condition mode feature vector, combined with a preset energy efficiency influence factor weight matrix, construct an energy efficiency influence factor set;

[0009] Perform working condition state clustering on the energy efficiency influence factor set based on the dynamic density clustering algorithm to generate an actuator operation mode classification tree;

[0010] Conduct energy consumption traceability analysis on the non-steady operation mode to generate an energy efficiency degradation feature map;

[0011] Input the energy efficiency degradation feature map into a pre-trained energy consumption optimization decision model, and output a dynamic regulation strategy set;

[0012] After the dynamic regulation strategy set takes effect, by comparing the deviation degree between the energy efficiency transition trajectory before and after the implementation of the strategy and the theoretical optimal energy consumption curve, iteratively update the weight distribution of the energy consumption optimization decision model, and output the final optimized control parameter package.

[0013] In a preferred embodiment, multi-source operation data streams of an electric actuator are collected, and time-series features are extracted from the multi-source operation data streams based on a sliding window to generate an actuator condition mode feature vector. Specifically:

[0014] Collect the motor current signal, valve position feedback signal, control instruction sequence, and vibration spectrum data of the electric actuator to form a multi-source heterogeneous operation data stream;

[0015] Adjust the size of the sliding window according to the change frequency of the control instruction sequence;

[0016] Extract time-series features from the multi-source heterogeneous operation data stream within the sliding window;

[0017] Perform dimensional splicing on the time-domain features, frequency-domain features, and wavelet transform features extracted within each sliding window to generate an actuator condition mode feature vector.

[0018] In a preferred embodiment, according to the actuator condition mode feature vector and in combination with a preset energy efficiency impact factor weight matrix, an energy efficiency impact factor set is constructed. Specifically:

[0019] Extract time-domain feature parameters, frequency-domain feature parameters, and wavelet transform feature parameters from the actuator condition mode feature vector;

[0020] Match the row vectors of the energy efficiency impact factor weight matrix with the corresponding type of feature parameters respectively;

[0021] Perform a weighted calculation operation on each feature parameter;

[0022] Accumulate the weighted calculation results according to the feature type to generate a multi-dimensional energy efficiency impact factor set.

[0023] In a preferred embodiment, based on the dynamic density clustering algorithm, the energy efficiency impact factor set is clustered by operating conditions to generate an actuator operation mode classification tree. Specifically:

[0024] Select the factors whose product of local density and relative distance in the multi-dimensional energy efficiency impact factor set exceeds the threshold as the initial clustering centers;

[0025] According to the distribution density of the initial clustering centers, divide the continuous operating conditions with similar density and continuous distribution of energy efficiency impact factors into the same clustering branches;

[0026] Classify the discrete factors that meet the density connectivity condition into adjacent clustering branches;

[0027] Construct a tree-like classification structure, and perform pruning optimization based on the mean and variance of the energy efficiency index of the clustering branches to generate an actuator operation mode classification tree including steady-state operation modes and non-steady-state operation modes.

[0028] In a preferred embodiment, an energy consumption traceability analysis is performed on the non-steady-state operation mode to generate an energy efficiency degradation characteristic map, specifically as follows:

[0029] Identify the operating condition state interval of the non-steady-state operation mode and determine the threshold range of the energy efficiency impact factor;

[0030] According to the threshold range of the energy efficiency impact factor, screen out the instruction segments with abnormal increase in energy consumption, extract the abnormal pulse patterns related to the control instruction sequence, and mark the start and end times of the abnormal pulse patterns;

[0031] Based on the start and end times of the abnormal pulse patterns, calculate the abnormal change amount of the motor current signal of the electric actuator and the lag response amount of the valve position feedback signal;

[0032] Generate an energy efficiency degradation characteristic map based on the numerical correlation relationship between the abnormal change amount of the motor current signal and the lag response amount of the valve position feedback signal.

[0033] In a preferred embodiment, input the energy efficiency degradation characteristic map into a pre-trained energy consumption optimization decision model to output a dynamic regulation strategy set, specifically as follows:

[0034] Construct a training set of the energy efficiency degradation characteristic map of the electric actuator with historical operation data, and train the energy consumption optimization decision model through supervised learning;

[0035] Normalize the characteristic parameters of the energy efficiency degradation characteristic map to obtain a standardized characteristic input matrix;

[0036] Input the standardized characteristic input matrix into the pre-trained energy consumption optimization decision model for feature matching and energy efficiency pattern recognition;

[0037] According to the results of feature matching and energy efficiency pattern recognition, output a dynamic regulation strategy set corresponding to the operating condition mode of the electric actuator.

[0038] In a preferred embodiment, after the dynamic regulation strategy set takes effect, by comparing the deviation degree between the energy efficiency transition trajectory before and after the implementation of the strategy and the theoretical optimal energy consumption curve, iteratively update the weight distribution of the energy consumption optimization decision model, and output the final optimized control parameter package, specifically as follows:

[0039] Real-time record the change trends of the motor current signal, valve position feedback signal, and vibration spectrum data of the electric actuator, and draw the actual energy efficiency transition trajectory;

[0040] Perform a point-by-point numerical comparison between the actual energy efficiency transition trajectory and the theoretical optimal energy consumption curve, and calculate the deviation degree between the actual energy efficiency transition trajectory and the theoretical optimal energy consumption curve;

[0041] The deviation degree is fed back as an error to the energy consumption optimization decision-making model, and the weight distribution inside the energy consumption optimization decision is iteratively adjusted based on the error backpropagation mechanism;

[0042] When the deviation degree drops to within the preset allowable range, the optimized final optimization control parameter package is output.

[0043] On the other hand, the present invention provides an energy consumption optimization system for an electric actuator based on big data analysis, including a feature extraction module, an influence factor construction module, a working condition state clustering module, an energy consumption traceability analysis module, a regulation strategy output module, and a model update module;

[0044] The feature extraction module collects multi-source operation data streams of the electric actuator, extracts time-series features from the multi-source operation data streams based on a sliding window, and generates an actuator working condition modal feature vector;

[0045] The influence factor construction module constructs an energy efficiency influence factor set according to the actuator working condition modal feature vector and in combination with a preset energy efficiency influence factor weight matrix;

[0046] The working condition state clustering module performs working condition state clustering on the energy efficiency influence factor set based on the dynamic density clustering algorithm to generate an actuator operation modal classification tree;

[0047] The energy consumption traceability analysis module performs energy consumption traceability analysis on the non-steady operation mode to generate an energy efficiency degradation feature map;

[0048] The regulation strategy output module inputs the energy efficiency degradation feature map into a pre-trained energy consumption optimization decision-making model and outputs a dynamic regulation strategy set;

[0049] After the dynamic regulation strategy set takes effect, the model update module iteratively updates the weight distribution of the energy consumption optimization decision-making model by comparing the deviation degree between the energy efficiency transition trajectory before and after the implementation of the strategy and the theoretical optimal energy consumption curve, and outputs the final optimization control parameter package.

[0050] The technical effects and advantages of the energy consumption optimization method and system for an electric actuator based on big data analysis according to the present invention:

[0051] 1. Real-time acquisition of multi-source operation data, and the use of the sliding window technology to extract time-series features to accurately reflect the working conditions of the electric actuator; in combination with a preset energy efficiency influence factor weight matrix, weighted accumulation of the extracted time-domain, frequency-domain, and wavelet transform features to construct a multi-dimensional energy efficiency influence factor set; using the dynamic density clustering algorithm to perform working condition state division on the energy efficiency influence factor set to form a classification tree including steady-state and non-steady operation modes; in the non-steady mode, identifying abnormal pulse patterns through energy consumption traceability analysis, and quantifying the coupling relationship between energy efficiency decay and response lag to generate an energy efficiency degradation feature map;

[0052] 2. Input the energy efficiency degradation characteristic map into the pre-trained decision model to output the dynamic regulation strategy. After the strategy takes effect, by comparing the deviation between the actual energy efficiency transition trajectory and the theoretical optimal curve, use error feedback to iteratively update the model weights and output the final optimized control parameter package. This effectively improves the operating efficiency of the electric actuator and the accuracy of energy consumption management, reduces equipment energy consumption and operating costs, enhances the adaptive ability, and provides a reliable technical guarantee for intelligent manufacturing and green energy conservation. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 Schematic diagram of the electric actuator energy consumption optimization method based on big data analysis of the present invention;

[0054] Figure 2 Schematic diagram of the structure of the electric actuator energy consumption optimization system based on big data analysis of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0055] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0056] Embodiment 1:

[0057] Figure 1 The electric actuator energy consumption optimization method based on big data analysis of the present invention is given, which includes the following steps:

[0058] Collect the multi-source operation data stream of the electric actuator, extract the time-series characteristics of the multi-source operation data stream based on a sliding window, and generate the actuator working condition modal feature vector;

[0059] According to the actuator working condition modal feature vector, combine the preset energy efficiency impact factor weight matrix to construct the energy efficiency impact factor set;

[0060] Based on the dynamic density clustering algorithm, perform working condition state clustering on the energy efficiency impact factor set to generate the actuator operation modal classification tree;

[0061] Conduct energy consumption traceability analysis on the non-steady operation mode to generate the energy efficiency degradation characteristic map;

[0062] Input the energy efficiency degradation characteristic map into the pre-trained energy consumption optimization decision model to output the dynamic regulation strategy set;

[0063] After the dynamic regulation strategy set takes effect, by comparing the deviation between the energy efficiency transition trajectory before and after the implementation of the strategy and the theoretical optimal energy consumption curve, the weight distribution of the energy consumption optimization decision model is iteratively updated, and the final optimized control parameter package is output.

[0064] Specifically, collect the multi-source operation data stream of the electric actuator, extract the time-series features of the multi-source operation data stream based on the sliding window, and generate the actuator working condition modal feature vector, including:

[0065] Collect the motor current signal, valve position feedback signal, control instruction sequence and vibration spectrum data of the electric actuator to form a multi-source heterogeneous operation data stream;

[0066] Among them, the motor current signal represents the magnitude of the current consumed by the driving motor of the electric actuator during operation, reflecting the load condition; the valve position feedback signal represents the actual feedback information for controlling the valve position in the electric actuator, reflecting the actual opening or closing degree; the control instruction sequence represents a series of instructions sent to the electric actuator, such as start, stop, speed regulation, etc., and these instructions usually carry timestamps; the vibration spectrum data refers to the spectrum data obtained by performing a fast Fourier transform on the vibration signal collected by the vibration sensor, which can be used to monitor abnormal vibrations or fault omens.

[0067] Exemplarily, assume that during a certain acquisition process: the motor current signal samples 1000 times per second to obtain 1000 data points within 1 second; the valve position feedback signal samples 100 times per second; the control instruction sequence may have only 3 to 5 instructions within a certain second; the vibration spectrum data is calculated by fast Fourier transform, and each calculation generates a vector containing 50 band amplitudes. After aligning these data according to their respective sampling times, a multi-source heterogeneous operation data stream is formed;

[0068] Adjust the size of the sliding window according to the change frequency of the control instruction sequence;

[0069] Specifically, the sliding window is used to divide the data within a continuous time period for local feature extraction. When the control instruction frequency is high (for example, the instruction is updated every 50 milliseconds), a shorter sliding window (for example, 200 milliseconds) is required to capture the rapidly changing information.

[0070] When the control instruction frequency is low (for example, the instruction is updated every 500 milliseconds), a longer sliding window (for example, 1 second or longer) can be used to fully extract the features.

[0071] Exemplarily, if the update frequency of the control instruction in a certain operation segment is once every 100 milliseconds, the sliding window can be set to 500 milliseconds, that is, each window contains approximately 5 control instructions; if the update frequency is once every 300 milliseconds, the window can be set to 1.5 seconds to ensure that each window contains sufficient data for feature extraction.

[0072] Perform time-series feature extraction on the multi-source heterogeneous operation data stream within the sliding window;

[0073] Specifically, for each sliding window, analyze the time-series data of each signal within the window and extract features that can describe the data change law, including:

[0074] Time-domain features: such as average value, standard deviation, maximum value, minimum value, etc.;

[0075] Frequency-domain features: such as main frequency components, spectral energy distribution, etc.;

[0076] Wavelet transform features: Extract multi-scale components through wavelet transform, such as the energy change of the signal at different scales.

[0077] Exemplarily, within a 500-millisecond window, the average value of the motor current signal can be calculated, and the main frequency component can be calculated by Fourier transform; for the vibration signal, wavelet transform is used to extract the detail coefficients and approximation coefficients.

[0078] Concatenate the time-domain features, frequency-domain features, and wavelet transform features extracted from each sliding window to generate the actuator condition mode feature vector;

[0079] Specifically, all the features extracted within each sliding window are combined in a certain order to form a high-dimensional feature vector, which can comprehensively reflect the operating state of the actuator within the window.

[0080] For example, the features extracted within a sliding window are: average value, standard deviation, maximum value, minimum value of the motor current; main frequency, spectral energy of the vibration signal; energies at different scales extracted by wavelet transform, etc.

[0081] Concatenate these features in the preset order to obtain the condition mode feature vector.

[0082] Exemplarily, assume the features extracted within a window are as follows:

[0083] Motor current signal: average value = 5A, standard deviation = 0.5A;

[0084] Valve position feedback signal: average value = 30°, standard deviation = 2°;

[0085] Vibration spectrum data: main frequency = 50Hz, spectral energy = 200;

[0086] Wavelet transform: Energy at scale 1 = 100, energy at scale 2 = 80.

[0087] Then the eigenvector of this window can be expressed as: [5, 0.5, 30, 2, 50, 200, 100, 80].

[0088] Specifically, according to the actuator operating condition modal eigenvector, combined with the preset energy efficiency impact factor weight matrix, an energy efficiency impact factor set is constructed, including:

[0089] Extract time-domain characteristic parameters, frequency-domain characteristic parameters, and wavelet transform characteristic parameters from the actuator operating condition modal eigenvector;

[0090] Specifically, the generated operating condition modal eigenvector contains information from different time domains and frequency domains. Separate different types of characteristic parameters in the vector and use them as time-domain, frequency-domain, and wavelet transform characteristic parameters respectively:

[0091] Time-domain characteristic parameters: such as average value, standard deviation, extreme value, etc.;

[0092] Frequency-domain characteristic parameters: such as main frequency, spectral energy;

[0093] Wavelet transform characteristic parameters: such as energy distribution at each scale, detail coefficients, etc.

[0094] Exemplarily, in the vector [5, 0.5, 30, 2, 50, 200, 100, 80], the first 4 numbers (5, 0.5, 30, 2) can be considered as time-domain characteristic parameters, the next 2 numbers (50, 200) as frequency-domain characteristic parameters, and the last 2 numbers (100, 80) as wavelet transform characteristic parameters

[0095] Match the row vectors of the energy efficiency impact factor weight matrix with the corresponding type of characteristic parameters respectively.

[0096] Specifically, in the preset energy efficiency impact factor weight matrix, each row corresponds to a type of characteristic parameter and assigns a weight to this characteristic parameter to reflect the importance of this characteristic to the energy efficiency impact.

[0097] For example, if the first row in the matrix corresponds to the average value of the motor current, its weight may be 0.3; the second row corresponds to the standard deviation, and its weight is 0.2, etc.

[0098] The matching process is to correspond each characteristic parameter extracted from the eigenvector with the weight in the corresponding row of the weight matrix.

[0099] Perform a weighted calculation operation on each characteristic parameter; specifically, multiply each extracted characteristic parameter by its corresponding weight to obtain the weighted value;

[0100] Specifically, according to the feature types (time domain, frequency domain, wavelet transform), the weighted values are summed up to generate a value reflecting the overall influence of this category of features, which is called the energy efficiency index.

[0101] The weighted calculation results are accumulated according to the feature types to generate a multi-dimensional energy efficiency influence factor set.

[0102] Specifically, based on the dynamic density clustering algorithm, the working condition states of the energy efficiency influence factor set are clustered to generate an actuator operation mode classification tree, including:

[0103] Select the factors whose product of local density and relative distance in the multi-dimensional energy efficiency influence factor set exceeds the threshold as the initial clustering centers;

[0104] Specifically, in the dynamic density clustering algorithm, the commonly used metrics include local density (i.e., the point density around a certain data point) and relative distance (i.e., the distance between this point and the point with a higher density).

[0105] For each factor point, the local density can be defined as the number of factor points within the radius.

[0106] For each factor point, the relative distance is expressed as the distance to the nearest point with a higher density than it.

[0107] If the product of the local density and the relative distance exceeds the preset threshold, then this factor point meets the requirements and is selected as the initial clustering center.

[0108] According to the distribution density of the initial clustering centers, the continuous working condition states with similar density and continuous distribution of energy efficiency influence factors are divided into the same clustering branches;

[0109] Specifically, for the selected initial clustering centers, according to their respective density values, the factors with close density and continuous distribution are grouped into the same branch, that is, these factors are considered to represent the same working condition state. For example, if the local densities of two clustering centers are both between 8 and 12 and the distance between them is small, they can be divided into the same clustering branch.

[0110] Suppose there are two initial clustering centers A and B in the energy efficiency influence factor set, with local densities of 9 and 10 respectively, and the relative distances are both within the allowable range, then A, B and the factors near them are grouped into the same clustering branch.

[0111] The discrete factors that meet the density connectivity condition are grouped into adjacent clustering branches;

[0112] Specifically, for those discrete factors not directly covered by the initial clustering centers, if they satisfy density connectivity with the factors within a certain clustering branch (i.e., they can be connected through a series of adjacent high-density points), these discrete factors are classified into that clustering branch. For example, if the distance between a discrete factor and a point within the clustering branch is relatively close and the local density of this discrete factor is high, it can be classified into the adjacent clustering branch according to the connectivity rule.

[0113] Construct a tree-like classification structure, and perform pruning optimization based on the mean and variance of the energy efficiency index of the clustering branches to generate an actuator operation mode classification tree that includes steady-state operation modes and non-steady-state operation modes.

[0114] Specifically, construct a tree-like structure according to the clustering results, with each node representing a clustering branch. Calculate the mean and variance of the energy efficiency index for each branch. Based on the mean and variance, pruning can be performed (i.e., removing noisy and overly scattered branches), thus forming a tree-like structure. The branches of the tree-like structure represent steady-state operation modes (small fluctuations, low variance) and non-steady-state operation modes (large fluctuations, high variance). Exemplarily, if the mean of the energy efficiency index within a certain clustering branch is 10 and the variance is 0.5, it can be considered that this branch corresponds to a steady state mode; while for another branch, the mean is 15 and the variance is 4, which corresponds to a non-steady state mode. Display these branches in a tree-like structure to form an actuator operation mode classification tree.

[0115] Specifically, conduct an energy consumption traceability analysis for non-steady-state operation modes to generate an energy efficiency degradation characteristic map, including:

[0116] Identify the operating condition state intervals of non-steady-state operation modes and determine the threshold range of energy efficiency impact factors;

[0117] Specifically, according to the operation mode classification tree, identify the operating condition state intervals representing non-steady-state operation (such as large vibrations and obvious fluctuations). Within these intervals, statistically analyze the distribution of energy efficiency impact factors and determine a threshold range, which is used to distinguish normal energy consumption from abnormal energy consumption. Exemplarily, assume that within the non-steady-state mode area, it is statistically obtained that the values of the energy efficiency impact factors are mainly distributed between 15 and 25, then the threshold range can be determined as 15 - 25.

[0118] According to the threshold range of energy efficiency impact factors, screen out the instruction segments with abnormal energy consumption increase, extract the abnormal pulse patterns related to the control instruction sequence, and mark the start and end times of the abnormal pulse patterns.

[0119] Specifically, within the threshold range of the energy efficiency impact factor, compare the energy consumption data corresponding to control instructions in different time periods, and filter out the instruction segments with sudden increases in energy consumption, which usually appear in the form of pulses. By marking the start and end times, an abnormal pulse pattern is formed. Exemplarily, if within a certain period of time, the energy efficiency impact factor suddenly rises from the normal value of 10 to above 20 and quickly returns after maintaining for 50 milliseconds, this period of time can be marked as an abnormal pulse pattern, and its start time and end time are recorded respectively.

[0120] Based on the start and end times of the abnormal pulse pattern, calculate the abnormal change amount of the motor current signal of the electric actuator and the lag response amount of the valve position feedback signal.

[0121] Specifically, according to the marked start and end times of the abnormal pulse pattern, extract the motor current signal and the valve position feedback signal in the corresponding time period, and calculate:

[0122] Abnormal change amount: It represents the difference between the motor current signal and the normal level during the abnormal pulse pattern.

[0123] Lag response amount: It represents the response delay of the valve position feedback signal to the control instruction, that is, the time difference between the signal change and the instruction issuance.

[0124] Based on the numerical correlation relationship between the abnormal change amount of the motor current signal and the lag response amount of the valve position feedback signal, generate an energy efficiency degradation characteristic map; specifically, calculate the correlation coefficient between the abnormal change amount and the lag response amount, analyze the correlation between the two, and construct a map based on this to reflect the energy efficiency degradation under the non-steady-state operation mode. The relationship between the two parameters can be represented by a scatter plot, a line chart, etc. If the two are positively correlated, it indicates that there is a strong coupling between abnormal energy consumption and response lag.

[0125] Specifically, input the energy efficiency degradation characteristic map into a pre-trained energy consumption optimization decision model, and output a dynamic regulation strategy set, including:

[0126] Construct a training set of the energy efficiency degradation characteristic map of the electric actuator with historical operation data, and train the energy consumption optimization decision model through supervised learning.

[0127] Specifically, use historical operation data to generate an energy efficiency degradation characteristic map and construct a training set as training samples.

[0128] Adopt a supervised learning method to train the energy consumption optimization decision model so that it can identify abnormal energy consumption situations according to the input degradation characteristic map.

[0129] During the training process, the input data and the pre-marked energy efficiency mode labels are used for model learning together.

[0130] Exemplarily, collect data within the past month, generate several energy efficiency degradation maps, and label which time periods are abnormal energy consumption patterns and which are normal patterns. Use these data to train a neural network model or a support vector machine model.

[0131] Normalize the characteristic parameters of the energy efficiency degradation feature map to obtain a standardized feature input matrix;

[0132] Input the standardized feature input matrix into a pre-trained energy consumption optimization decision model for feature matching and energy efficiency pattern recognition;

[0133] Specifically, the standardized feature input matrix is used as the input of the energy consumption optimization decision model. After training, the model can automatically match the relationship between the input features and the historical energy efficiency patterns and identify the current energy efficiency state.

[0134] The model may adopt algorithms such as deep neural networks and random forests to automatically output prediction results based on the input data.

[0135] Exemplarily, after inputting the normalized feature vector into the model, the output result of the model may be steady-state operation or non-steady-state operation, and corresponding regulation suggestions are given.

[0136] Output a dynamic regulation strategy set corresponding to the operating condition mode of the electric actuator according to the feature matching and energy efficiency pattern recognition results;

[0137] Specifically, the energy consumption optimization decision model generates a dynamic regulation strategy for the electric actuator according to the identified energy efficiency pattern, including adjusting the motor current, changing the valve position, or modifying the control instruction frequency, etc.

[0138] The output dynamic regulation strategy set is the combination of regulation parameters that best matches the current operating condition mode.

[0139] Exemplarily, if the model identifies that the current is in a non-steady-state operation state, the possible output strategies may be to reduce the current setting value by 0.5 A, extend the valve response time by 10 ms, etc.

[0140] Specifically, after the dynamic regulation strategy set takes effect, by comparing the deviation degree between the energy efficiency transition trajectory before and after the implementation of the strategy and the theoretical optimal energy consumption curve, iteratively update the weight distribution of the energy consumption optimization decision model, and output the final optimized control parameter package, including:

[0141] Real-time record the change trends of the motor current signal, valve position feedback signal, and vibration spectrum data of the electric actuator, and draw the actual energy efficiency transition trajectory;

[0142] Specifically, after the dynamic regulation strategy takes effect, the system continuously collects the key operation data of the actuator and draws a trajectory diagram reflecting the energy efficiency change based on these data.

[0143] The trajectory graph reflects the change trend of energy consumption over time and can show the changes before and after the implementation of the regulation strategy.

[0144] Exemplarily, before regulation, the energy efficiency transition trajectory may show large fluctuations; after regulation, the trajectory tends to be stable and closer to the theoretical optimal curve.

[0145] Perform a point-by-point numerical comparison between the actual energy efficiency transition trajectory and the theoretical optimal energy consumption curve, and calculate the deviation degree between the actual energy efficiency transition trajectory and the theoretical optimal energy consumption curve;

[0146] Specifically, compare the actually collected energy efficiency data with the preset theoretical optimal energy consumption curve, calculate the difference between the two at each time point, and then obtain the overall deviation degree. Exemplarily, if at a certain time point, the actual energy efficiency value is 8 and the theoretical value is 7, then the error at this point is 8 - 7 = 1; after calculating for all time points, take the mean square error value as the overall deviation degree.

[0147] Feed the deviation degree back to the energy consumption optimization decision model as an error, and iteratively adjust the weight distribution inside the energy consumption optimization decision based on the error backpropagation mechanism;

[0148] Specifically, use the error backpropagation algorithm, take the calculated deviation degree (such as the mean square error value) as the output of the loss function, and feed it back to the pre-trained energy consumption optimization decision model, so as to adjust the weights of each layer inside the model and make the next prediction more accurate.

[0149] When the deviation degree drops to the preset allowable range, output the optimized final optimized control parameter package;

[0150] Specifically, after iterative adjustment, when the deviation degree between the actual energy efficiency transition trajectory and the theoretical optimal energy consumption curve is lower than the preset allowable threshold (for example, the mean square error is lower than 0.1), it is considered that the model has reached the optimal state. At this time, output the final control parameter package to guide the subsequent operation mode switching of the electric actuator.

[0151] Exemplarily, if the preset allowable mean square error threshold is 0.1, when the model is iterated and the measured mean square error is 0.08, then output the final optimized control parameter package, which includes parameters such as the adjusted motor current setting value and valve action time.

[0152] Example 2:

[0153] The difference between Embodiment 2 and Embodiment 1 of the present invention is that this embodiment introduces an energy consumption optimization system for an electric actuator based on big data analysis.

[0154] Figure 2Fig. 0 shows a schematic structural diagram of the energy consumption optimization system for electric actuators based on big data analysis. The energy consumption optimization system for electric actuators based on big data analysis includes a feature extraction module, an influence factor construction module, a working condition state clustering module, an energy consumption traceability analysis module, a regulation strategy output module, and a model update module;

[0155] The feature extraction module collects multi-source operation data streams of the electric actuator, extracts time-series features from the multi-source operation data streams based on a sliding window, and generates an actuator working condition mode feature vector;

[0156] The influence factor construction module constructs an energy efficiency influence factor set according to the actuator working condition mode feature vector and in combination with a preset energy efficiency influence factor weight matrix;

[0157] The working condition state clustering module performs working condition state clustering on the energy efficiency influence factor set based on the dynamic density clustering algorithm, and generates an actuator operation mode classification tree;

[0158] The energy consumption traceability analysis module performs energy consumption traceability analysis on the non-steady operation mode, and generates an energy efficiency degradation feature map;

[0159] The regulation strategy output module inputs the energy efficiency degradation feature map into a pre-trained energy consumption optimization decision model, and outputs a dynamic regulation strategy set;

[0160] After the dynamic regulation strategy set takes effect, the model update module iteratively updates the weight distribution of the energy consumption optimization decision model by comparing the deviation degree between the energy efficiency transition trajectory before and after the strategy implementation and the theoretical optimal energy consumption curve, and outputs a final optimized control parameter package.

[0161] All the above formulas are dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the actual situation. The preset parameters and threshold selection in the formulas are set by those skilled in the art according to the actual situation.

[0162] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that contains one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, a hard disk, or a magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0163] Those of ordinary skill in the art will realize that the modules and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. A professional technician can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0164] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and modules described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0165] In several embodiments provided in the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules is only a logical function division, and there can be other division methods in actual implementation. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings, direct couplings, or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the devices or modules can be in electrical, mechanical, or other forms.

[0166] The module described as a separation component may or may not be physically separated. The component shown as a module may or may not be a physical module, and it may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0167] In addition, in each embodiment of this application, each functional module can be integrated into one processing module, or each module can exist physically alone, or two or more modules can be integrated into one module.

[0168] If the described function is implemented in the form of a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of this application. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0169] As described above, this is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed in this application can easily think of changes or substitutions, which should all be covered by the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claimed rights.

[0170] Finally: The above is only the preferred embodiment of the present invention and is not used to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An energy consumption optimization method for an electric actuator based on big data analysis, characterized in that, It includes the following steps: Collect the multi-source operation data stream of the electric actuator, extract the time-series features of the multi-source operation data stream based on a sliding window, and generate the actuator condition mode feature vector; According to the actuator condition mode feature vector, combined with the preset energy efficiency impact factor weight matrix, construct the energy efficiency impact factor set; Based on the dynamic density clustering algorithm, perform condition state clustering on the energy efficiency impact factor set to generate the actuator operation mode classification tree; Select the factors whose product of local density and relative distance in the multi-dimensional energy efficiency impact factor set exceeds the threshold as the initial clustering centers; According to the distribution density of the initial clustering centers, divide the continuous condition states with similar density and continuous distribution of energy efficiency impact factors into the same clustering branches; Classify the discrete factors that meet the density connectivity condition into adjacent clustering branches; Construct a tree-like classification structure, perform pruning optimization based on the mean and variance of the energy efficiency index of the clustering branches, and generate the actuator operation mode classification tree including the steady-state operation mode and the non-steady-state operation mode; Conduct energy consumption traceability analysis on the non-steady-state operation mode to generate the energy efficiency degradation feature map; Identify the condition state interval of the non-steady-state operation mode and determine the threshold range of the energy efficiency impact factors; According to the threshold range of the energy efficiency impact factors, screen the instruction segments with abnormal increase in energy consumption, extract the abnormal pulse patterns related to the control instruction sequence, and mark the start and end times of the abnormal pulse patterns; Based on the start and end times of the abnormal pulse patterns, calculate the abnormal change amount of the motor current signal and the lag response amount of the valve position feedback signal of the electric actuator; Based on the numerical correlation relationship between the abnormal change amount of the motor current signal and the lag response amount of the valve position feedback signal, generate the energy efficiency degradation feature map; Input the energy efficiency degradation feature map into the pre-trained energy consumption optimization decision model to output the dynamic regulation strategy set; After the dynamic regulation strategy set takes effect, by comparing the deviation degree between the energy efficiency transition trajectory before and after the implementation of the strategy and the theoretical optimal energy consumption curve, iteratively update the weight distribution of the energy consumption optimization decision model and output the final optimized control parameter package.

2. The energy consumption optimization method of the electric actuator based on big data analysis according to claim 1, wherein Collect the multi-source operation data stream of the electric actuator, extract the time-series features of the multi-source operation data stream based on a sliding window, and generate the actuator condition mode feature vector. Specifically: Collect the motor current signal, valve position feedback signal, control instruction sequence, and vibration spectrum data of the electric actuator to form a multi-source heterogeneous operation data stream; Adjust the size of the sliding window according to the change frequency of the control instruction sequence; Extract the time-series features of the multi-source heterogeneous operation data stream within the sliding window; Perform dimension splicing on the time-domain features, frequency-domain features, and wavelet transform features extracted within each sliding window to generate the actuator condition mode feature vector.

3. The energy consumption optimization method of the electric actuator based on big data analysis according to claim 2, wherein, According to the actuator condition mode feature vector, combined with the preset energy efficiency impact factor weight matrix, construct the energy efficiency impact factor set. Specifically: Extract the time-domain feature parameters, frequency-domain feature parameters, and wavelet transform feature parameters from the actuator condition mode feature vector; Match the row vectors of the energy efficiency impact factor weight matrix with the corresponding type of feature parameters respectively; Perform weighted calculation operations on each feature parameter; Accumulate the weighted calculation results according to the feature type to generate the multi-dimensional energy efficiency impact factor set.

4. The energy consumption optimization method of the electric actuator based on big data analysis according to claim 3, characterized in that Input the energy efficiency degradation feature map into the pre-trained energy consumption optimization decision model to output a set of dynamic regulation strategies, specifically: Construct a training set of the energy efficiency degradation feature map of the electric actuator with historical operation data, and train the energy consumption optimization decision model through supervised learning; Normalize the characteristic parameters of the energy efficiency degradation feature map to obtain a standardized characteristic input matrix; Input the standardized characteristic input matrix into the pre-trained energy consumption optimization decision model for feature matching and energy efficiency mode recognition; According to the results of feature matching and energy efficiency mode recognition, output a set of dynamic regulation strategies corresponding to the operation condition mode of the electric actuator.

5. The energy consumption optimization method of the electric actuator based on big data analysis according to claim 4, characterized in that After the set of dynamic regulation strategies takes effect, by comparing the deviation degree between the energy efficiency transition trajectory before and after the implementation of the strategy and the theoretical optimal energy consumption curve, iteratively update the weight distribution of the energy consumption optimization decision model to output the final optimized control parameter package, specifically: Record the change trends of the motor current signal, valve position feedback signal, and vibration spectrum data of the electric actuator in real time, and draw the actual energy efficiency transition trajectory; Make a point-by-point numerical comparison between the actual energy efficiency transition trajectory and the theoretical optimal energy consumption curve, and calculate the deviation degree between the actual energy efficiency transition trajectory and the theoretical optimal energy consumption curve; Feed the deviation degree back to the energy consumption optimization decision model as an error, and iteratively adjust the weight distribution inside the energy consumption optimization decision based on the error backpropagation mechanism; When the deviation degree drops to the preset allowable range, output the optimized final optimized control parameter package.

6. An electric actuator energy consumption optimization system based on big data analysis, which is used to implement the electric actuator energy consumption optimization method based on big data analysis according to any one of claims 1-5, and is characterized in that, It includes a feature extraction module, an influencing factor construction module, a working condition state clustering module, an energy consumption traceability analysis module, a regulation strategy output module, and a model update module; The feature extraction module collects the multi-source operation data stream of the electric actuator, extracts the time-series features of the multi-source operation data stream based on a sliding window, and generates the actuator working condition mode feature vector; The influencing factor construction module constructs a set of energy efficiency influencing factors according to the actuator working condition mode feature vector and in combination with the preset energy efficiency influencing factor weight matrix; The working condition state clustering module performs working condition state clustering on the set of energy efficiency influencing factors based on the dynamic density clustering algorithm to generate an actuator operation mode classification tree; Select the factors whose product of local density and relative distance in the multi-dimensional energy efficiency influencing factor set exceeds the threshold as the initial clustering centers; According to the distribution density of the initial clustering centers, divide the continuous working condition states with similar density and continuous distribution of energy efficiency influencing factors into the same clustering branch; Classify the discrete factors that meet the density connectivity condition into adjacent clustering branches; Construct a tree-shaped classification structure, and perform pruning optimization based on the mean and variance of the energy efficiency index of the clustering branches to generate an actuator operation mode classification tree including steady-state operation modes and non-steady-state operation modes; The energy consumption traceability analysis module performs energy consumption traceability analysis on the non-steady-state operation mode to generate an energy efficiency degradation feature map; Identify the working condition state interval of the non-steady-state operation mode and determine the threshold range of the energy efficiency influencing factors; According to the threshold range of the energy efficiency influencing factors, screen the instruction segments with abnormal increase in energy consumption, extract the abnormal pulse patterns related to the control instruction sequence, and mark the start and end times of the abnormal pulse patterns; Based on the start and end times of the abnormal pulse mode, calculate the abnormal change amount of the motor current signal of the electric actuator and the lag response amount of the valve position feedback signal; Generate an energy efficiency degradation characteristic map based on the numerical correlation relationship between the abnormal change amount of the motor current signal and the lag response amount of the valve position feedback signal; The regulation strategy output module inputs the energy efficiency degradation characteristic map into the pre-trained energy consumption optimization decision model and outputs a set of dynamic regulation strategies; After the set of dynamic regulation strategies takes effect, the model update module iteratively updates the weight distribution of the energy consumption optimization decision model by comparing the deviation degree between the energy efficiency transition trajectory before and after the implementation of the strategy and the theoretical optimal energy consumption curve, and outputs the final optimized control parameter package.

Images