Energy-saving electro-hydraulic driving material distribution trolley and hydraulic integrated module control method

Abstract

The present application relates to the technical field of hydraulic integrated module control using models, and particularly relates to an energy-saving electro-hydraulic driving material distribution trolley and a hydraulic integrated module control method. The present application first analyzes the distribution concentration of electro-hydraulic driving parameters in clustering based on state vectors to obtain important coefficients; further analyzes the matching degree of samples in two types of clusters of each type of state vector in limited state vector categories and unrestricted clustering based on electro-hydraulic driving parameters to obtain important correction coefficients; further obtains the influence degree according to the change trend of the initial training weight; further corrects the initial training weight to obtain the correction training weight of the next training according to the training error of the current training, in combination with the influence degree and important coefficients of the samples, until the model converges, so that more accurate and optimal electro-hydraulic driving parameters of the trolley can be provided in the prediction process, and the energy-saving effect is guaranteed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of hydraulic integrated module control using models, in particular to an energy-saving electro-hydraulic drive material distribution trolley and a hydraulic integrated module control method. BACKGROUND

[0002] Electro-hydraulic drive can enable the material distribution trolley to work stably in harsh environments with heavy load, dust and strong vibration, and maintenance is relatively simple, prolonging the service life. The hydraulic module is the power core of the material distribution trolley, which can adjust the flow, direction and other information of the hydraulic oil through the control valve group, realize precise control of the flexibility of the movement direction, and the existing machine learning method trains the historical operation data and related parameters of the material distribution trolley to obtain the best parameters for electro-hydraulic drive.

[0003] Since the historical data of the trolley is affected by the running state of the trolley, the trolley does not run according to its own best parameters, and there may be a large difference in the corresponding electro-hydraulic parameters of similar trolley working states, thereby making it difficult to determine the best parameters of the trolley. The obtained parameters have a large difference from the best parameters that should be adopted by the trolley, thereby affecting the accuracy of the electro-hydraulic drive control of the trolley, and thus causing energy waste in the use of the material distribution trolley and failing to achieve the purpose of energy saving. SUMMARY

[0004] In order to solve the technical problem that the historical data is affected by the running state of the trolley, the control accuracy is insufficient by machine learning, and the energy saving effect is affected, the purpose of the present application is to provide an energy-saving electro-hydraulic drive material distribution trolley and a hydraulic integrated module control method, and the technical solution adopted is as follows:

[0005] A hydraulic integrated module control method of an energy-saving electro-hydraulic drive material distribution trolley, the method comprising:

[0006] Obtaining historical material distribution data samples of the trolley, including a state vector and electro-hydraulic drive parameters;

[0007] In the clustering based on the state vector of the sample, according to the distribution concentration of the electro-hydraulic drive parameters in each cluster, the important coefficient of each sample is obtained; in each type of state vector sample, clustering is performed based on the electro-hydraulic drive parameters to obtain a first type of cluster; clustering is performed on all samples based on the electro-hydraulic drive parameters to obtain a second type of cluster; the matching degree of each type of state vector sample corresponding to the first type of cluster and the second type of cluster is analyzed, and the important correction coefficient of each sample is obtained; the initial training weight is obtained by fusing the important coefficient and the important correction coefficient;

[0008] According to the change trend of the initial training weight corresponding to the same state vector of the same trolley, the influence degree of each sample is obtained; according to the training error of the current training, the initial training weight is corrected to obtain the corrected training weight of each sample in the next training process in combination with the influence degree and the important coefficient of the sample, and the model converges until the model converges;

[0009] The state vector of the current trolley is input into the trained model to obtain the electro-hydraulic driving parameter.

[0010] Further, the method for obtaining the important coefficient comprises:

[0011] For each cluster: by sorting the number of each type of the electro-hydraulic driving parameter, the important coefficient of the sample in each cluster is obtained according to the difference in the proportion of the number of samples of the two types of the electro-hydraulic driving parameter with the largest number and the second largest number.

[0012] Further, the method for obtaining the important correction coefficient comprises:

[0013] In each type of the state vector sample: according to the cluster matching of the proportion of the common samples in the first type of cluster and the second type of cluster; according to the number similarity of the matched second type of cluster and the first type of cluster, the important correction coefficient of each sample is obtained in combination with the proportion of the common samples in the matching.

[0014] Further, the method for obtaining the influence degree comprises:

[0015] In time sequence, the initial training weights of all samples of the same trolley are sorted to construct a weight sequence, the initial training weights corresponding to each same state vector are extracted and the order in the weight sequence is retained to construct a subsequence;

[0016] The slope of the fitting straight line of each subsequence is extracted to obtain the influence factor of each subsequence; according to the overall characteristics of all the influence factors of all trolleys, the influence degree of each sample is obtained in combination with the influence factor of the subsequence to which each sample belongs.

[0017] Further, the method for obtaining the corrected training weight of each sample in the next training process comprises:

[0018] During the first training, the initial training weight is taken as the corrected training weight; the predicted parameter of the electro-hydraulic driving of each sample is obtained based on the corrected training weight of the current training;

[0019] The predicted parameter is corrected according to the influence degree and the important coefficient, and the training error of each sample in the current training is obtained according to the difference between the corrected predicted parameter and the electro-hydraulic driving parameter of the sample.

[0020] Fuse the training error and the initial training weight, and obtain the correction training weight of each sample in the next training process.

[0021] Further, the determination method of the model convergence comprises:

[0022] Further, the determination method of the model convergence comprises:

[0023] Further, the determination method of the model convergence comprises:

[0024] Further, the determination method of the model convergence comprises:

[0025] Further, the determination method of the model convergence comprises:

[0026] Further, the determination method of the model convergence comprises:

[0027] Further, the determination method of the model convergence comprises:

[0028] The application further provides an energy-saving electro-hydraulic driving material distributing trolley, which comprises a trolley body and a hydraulic integrated module, and further comprises a system comprising a memory, a processor and a computer program stored in the memory and capable of running on the processor, and the processor implements the steps of the hydraulic integrated module control method of any one of the energy-saving electro-hydraulic driving material distributing trolleys.

[0029] The application has the following beneficial effects:

[0030] The application firstly acquires historical distribution data samples of the trolley, and analyzes the distribution concentration of electro-hydraulic driving parameters in clustering based on state vectors, acquires important coefficients, and preliminarily evaluates the degree of attention of the samples, so as to provide a basis for acquiring initial training weights; further, in the clustering of limiting state vector categories and unrestricted electro-hydraulic driving parameters, the matching degree of samples in two types of clusters of each state vector is analyzed, important correction coefficients are acquired, the model is prevented from learning only the average characteristics of most samples and ignoring special trolleys that still show parameter differences under similar working conditions, and prediction deviation is reduced; further, according to the change trend of the initial training weights corresponding to the same state vector of the same trolley, the influence of the performance change of the trolley itself is analyzed, the influence degree is acquired, and a basis is provided for more accurate analysis of training errors and adjustment of training weights in the future; further, according to the training error of the current training, the influence degree and important coefficients of the samples are combined to correct the initial training weights to obtain the correction training weights of the next training, until the model converges, the problem that historical data are affected by the running state of the trolley, resulting in an unsatisfactory machine learning result and inaccurate prediction control is solved, and more accurate and optimal electro-hydraulic driving parameters of the trolley can be provided in the prediction process, so as to guarantee the energy-saving effect. BRIEF DESCRIPTION OF DRAWINGS

[0031] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, and the advantages thereof, a brief introduction will be given to the drawings needed in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor based on these drawings.

[0032] Figure 1 A flow chart of a hydraulic integrated module control method of an energy-saving electro-hydraulic driving distribution trolley provided by an embodiment of the present application is shown in the figure.

[0033] Figure 2 A flow chart of an influence degree acquisition method provided by an embodiment of the present application is shown in the figure.

[0034] Figure 3 A flow chart of an acquisition method of correction training weights of a next training process provided by an embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0035] In order to further clarify the technical means and effects taken by the present application to achieve the predetermined inventive purpose, the specific implementation, structure, features and effects of the energy-saving electro-hydraulic drive material distribution trolley and hydraulic integrated module control method according to the present application are described in detail as follows. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0036] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs.

[0037] The specific scheme of the energy-saving electro-hydraulic drive material distribution trolley and hydraulic integrated module control method provided by the present application is described in detail below in conjunction with the drawings.

[0038] Please refer to Figure 1 which shows a flow chart of the hydraulic integrated module control method of an energy-saving electro-hydraulic drive material distribution trolley according to an embodiment of the present application, which specifically includes:

[0039] Step S1: Obtain historical material distribution data samples of the trolley, including state vectors and electro-hydraulic drive parameters.

[0040] In an embodiment of the present application, the working condition data of the trolley in each material distribution is obtained through various sensors, including at least: material weight, trolley speed, hydraulic system pressure, vibration and load current; and the corresponding electro-hydraulic drive parameters, specifically output current, are obtained, which are time series data.

[0041] The state vector of the working condition data of each sample is obtained by using LeNet feature extraction network, so as to establish the "state vector-electro-hydraulic drive parameter" data sample of each material distribution; and the Transformer network is constructed, the state vector is the input data, and the corresponding electro-hydraulic drive parameter is output.

[0042] It should be noted that the historical material distribution data samples can be limited to the historical data of the last year, and the data collection frequency, collection method and time range of the historical data can be set by the implementer according to actual needs; in other embodiments of the present application, an autoencoder can also be used to construct an encoder-decoder structure to compress the original data through unsupervised learning, and the hidden layer output of the encoder is the state vector, which is a known technology as LeNet feature extraction network and Transformer network, and will not be described here.

[0043] Step S2: In the clusters where samples are clustered based on state vectors, obtain the importance coefficient of each sample according to the distribution concentration of electro-hydraulic drive parameters within each cluster; in the samples of each class of state vectors, cluster based on electro-hydraulic drive parameters to obtain the first class of clusters; cluster all samples based on electro-hydraulic drive parameters to obtain the second class of clusters; analyze the matching degree of the samples in the first and second classes of clusters corresponding to the samples of each class of state vectors to obtain the importance correction coefficient of each sample; fuse the importance coefficient and the importance correction coefficient to obtain the initial training weights.

[0044] In clustering based on state vectors, vehicle data samples with similar working states are grouped into one class. Since changes in electro-hydraulic drive parameters directly affect the working state of the vehicle, the electro-hydraulic drive parameters of data samples with similar working states should also be concentrated. Therefore, the concentration of electro-hydraulic drive parameters within each cluster reflects the normal probability of the electro-hydraulic drive parameters, thereby obtaining the importance coefficient of each sample, initially assessing the degree of focus of the samples, and providing a basis for obtaining the initial training weights.

[0045] Preferably, in one embodiment of the present invention, in a cluster that clusters samples based on state vectors, for each cluster: firstly, each type of electro-hydraulic driving parameter can be regarded as a sub-cluster, and sorted by the number of corresponding electro-hydraulic driving parameters of each type to obtain the sorting sequence of all sub-clusters in a cluster;

[0046] Considering that the proportion of samples of the most numerous electro-hydraulic driving parameters is higher, while the proportion of samples of the second most numerous electro-hydraulic driving parameters is lower, it indicates that the electro-hydraulic driving parameters are concentrated in a sub-cluster. The stronger the concentration of distribution, the higher the accuracy of judging electro-hydraulic driving parameters under similar state vectors, and thus the smaller the sample specificity, and the smaller its importance in subsequent analysis.

[0047] Based on this, the importance coefficient of samples within each cluster is obtained by considering the difference in the proportion of samples with the most and second most combined electro-hydraulic drive parameters.

[0048] In clusters that cluster samples based on state vectors, for each cluster: the proportion of the number of samples of a certain type of electro-hydraulic driving parameter to the total number of samples in the entire cluster is taken as the quantity ratio; the difference between the quantity ratio of the most numerous electro-hydraulic driving parameter and the quantity ratio of the second most numerous electro-hydraulic driving parameter is taken as the numerator, and the quantity ratio of the second most numerous electro-hydraulic driving parameter is taken as the denominator. After linearly normalizing the fractional ratio as the independent variable in the corresponding data dimension, the difference between the constant 1 and the normalization result is taken as the importance coefficient of each sample in each cluster.

[0049] Wherein, the clustering is performed by the ISODATA algorithm, the same electro-hydraulic driving parameters are a class, the number distribution of the samples of each electro-hydraulic driving parameter in a cluster is expressed by the number ratio, the number ratio difference between the two classes of electro-hydraulic driving parameters with the largest number and the second largest number is expressed by measuring the relative deviation, so as to reflect the distribution concentration of the electro-hydraulic driving parameters in the cluster, the larger the normalized result is, the stronger the distribution concentration is, and then the negative correlation mapping is performed by means of the constant 1, to obtain the importance coefficient of each sample in the cluster; the analysis mode of each cluster based on the state vector is consistent, and only one example is described here, and the description is not repeated.

[0050] In other embodiments of the application, the implementer can also perform negative correlation normalization mapping by using the exp(-x) function with the natural constant e as the base, taking the fractional ratio to be mapped as the function independent variable, and taking the function mapping result as the importance coefficient; x is the independent variable, and the ISODATA algorithm is a known technology and will not be described here.

[0051] However, due to factors such as the service life of the trolley, the transportation load and other factors, the trolley working state under similar working conditions adopts obviously different electro-hydraulic driving parameters, which indicates that small differences in the state characteristics of the trolley can lead to large differences in the electro-hydraulic driving parameters, and further lead to inaccurate prediction of the electro-hydraulic driving parameters of the trolley. Therefore, it is necessary to analyze the differences in the electro-hydraulic driving parameters under similar state vectors to further evaluate the importance of the samples.

[0052] Considering that the samples in each class of state vectors are clustered based on the electro-hydraulic driving parameters to obtain the first type of cluster, which can reflect the internal difference distribution of the parameters under the limited working condition; at the same time, all samples are clustered based on the electro-hydraulic driving parameters to obtain the second type of cluster, which reflects the overall distribution pattern of the parameters under different working conditions; then the matching degree of the samples in the first type of cluster and the second type of cluster corresponding to each class of state vectors is analyzed, which can reveal whether there are differences in driving parameters between different trolleys under the same or similar working conditions. Such differences often come from individual factors such as service life, load condition, wear degree, etc. By comparing the local and global, the importance correction coefficient of each sample is obtained, avoiding the model to learn only the average characteristics of most samples and ignoring the special trolley that still shows parameter difference under similar working conditions, reducing the prediction deviation and improving the prediction accuracy.

[0053] Preferably, in one embodiment of the application, the same state vector is a class; in the samples of each class of state vectors: considering that the more the proportion of common samples between different clusters is, the higher the matching degree is, so first, the cluster matching is performed according to the proportion of common samples in the first type of cluster and the second type of cluster;

[0054] As an example, the matching degree of the samples in the two clusters is represented by the Jaccard index of the two clusters, the Jaccard index representing the proportion of the common samples in the first type of cluster and the second type of cluster, the first type of cluster is matched with the second type of cluster corresponding to the maximum Jaccard index of the first type of cluster; the Jaccard index of each first type of cluster and each second type of cluster is calculated one by one, the matched second type of cluster of each first type of cluster is obtained, and different first type of clusters can simultaneously match the same second type of cluster.

[0055] The Jaccard index refers to the intersection of the samples in the two clusters as the numerator and the union as the denominator.

[0056] It is considered that the more the matched second type of cluster, the closer to the number of the first type of cluster, which indicates that under the same state vector, there is an obvious difference in the electro-hydraulic driving parameter, and the corresponding sample needs to be focused on, and the important correction coefficient is higher; meanwhile, the higher the proportion of the common sample in the matching, the higher the matching degree, and the higher the reliability of the matching relationship, based on which, the important correction coefficient of each sample is obtained according to the number similarity of the matched second type of cluster and the first type of cluster and the proportion of the common sample in the matching.

[0057] In the samples of each type of state vector: the number of the matched second type of cluster is taken as the numerator, the number of the first type of cluster is taken as the denominator, and the fractional ratio is taken as the number similarity parameter, representing the number similarity of the matched second type of cluster and the first type of cluster; the product of the proportion of the common sample (the maximum Jaccard index) in the matching of each first type of cluster and the number similarity parameter is taken as the important correction coefficient of each sample in each first type of cluster.

[0058] The analysis mode of the samples of each type of state vector is consistent, and only one example is described here, and no repeated description is given.

[0059] It should be noted that before the electro-hydraulic driving parameters are clustered, the DTW distances of the electro-hydraulic driving parameters of different samples are calculated to obtain a DTW distance matrix, and then the ISODATA algorithm is used for clustering; in other embodiments of the present application, hierarchical clustering can also be used to cluster the state vectors, the DTW distance is used to replace the Euclidean distance, and K-Means is used to cluster the electro-hydraulic driving parameters, which are all prior art and will not be described here.

[0060] The important coefficient reflects the particularity of the sample in the overall distribution under similar state vectors, and can reflect the importance of the state characteristics; the important correction coefficient is used to correct the deviation of the driving parameter caused by the individual difference of the trolley, and reflects the sensitivity and reference value of the sample to the model training under similar working conditions, so the important coefficient and the important correction coefficient are fused, and at the same time, in order to speed up the training process of the neural network, the fusion result is taken as the initial training weight of the sample.

[0061] Preferably, in one embodiment of the present application, the product of the important coefficient of the sample and the important correction coefficient is taken as the initial training weight.

[0062] In other embodiments of the present application, the average value or weighted sum result of the important coefficient of the sample and the important correction coefficient can also be taken as the initial training weight.

[0063] Step S3: According to the change trend of the initial training weight corresponding to the same state vector of the same trolley, the influence degree of each sample is obtained; according to the training error of the current training, the initial training weight is corrected to obtain the correction training weight of each sample in the next training process in combination with the influence degree and the important coefficient of the sample, until the model converges.

[0064] Since the use time of different trolleys is different, for example, some trolleys may just start to be put into use, and some trolleys have been in use for a long time; and the same state vector of the same trolley represents the same working condition state of the trolley, when the trolley is used for many times, the performance of the trolley itself may be reduced, and the electro-hydraulic driving parameters corresponding to the trolley also change, so the initial training weight corresponding to the change trend of the initial training weight corresponding to the same state vector of the same trolley also changes, so the influence degree of each sample is obtained according to the change trend of the initial training weight corresponding to the same state vector of the same trolley, which provides a basis for more accurately analyzing the training error and adjusting the training weight.

[0065] Preferably, in one embodiment of the present application, please refer to Figure 2 which shows a flowchart of an influence degree acquisition method provided by one embodiment of the present application, and specifically includes:

[0066] Step S301: The initial training weights of all samples of the same trolley are sorted in time sequence to construct a weight sequence, the initial training weights corresponding to each type of the same state vector are extracted and the order in the weight sequence is preserved to construct a subsequence.

[0067] In one embodiment of the present application, in order to analyze the change of the initial training weight, the time axis is taken as the horizontal axis, the data axis of the initial training weight is taken as the vertical axis, the termination time point of each sample is taken as the horizontal axis coordinate, and the initial training weight is taken as the vertical axis coordinate, the initial training weights of all samples of the same trolley are mapped into a two-dimensional coordinate system in time sequence, and a weight sequence is obtained.

[0068] Selecting any type of state vector, the data points of other state vectors are excluded from the coordinate system, and only the data points of the selected type of state vector are retained to obtain a subsequence.

[0069] The analysis method of each type of state vector of each trolley and the analysis method of the influence degree of each sample are consistent, and only one example is described here, and no repeated description is made.

[0070] Step S302: Extract the slope of the fitting straight line of each sub-sequence to obtain the influence factor of each sub-sequence; according to the overall characteristics of all influence factors of all trolleys, combined with the influence factor of the sub-sequence to which each sample belongs, the influence degree of each sample is obtained.

[0071] Considering that the trolley will reduce its performance after multiple uses, it needs to provide larger electro-hydraulic driving parameters to meet the material distribution or transportation under the same condition, at this time the electro-hydraulic driving parameter distribution is more concentrated, and when the important correction coefficient is matched, more second type clusters are matched, resulting in the initial training weight gradually increasing, so the slope of the fitting straight line of each sub-sequence is extracted to obtain the influence factor of each sub-sequence.

[0072] Considering that the overall characteristics of all influence factors of all sub-sequences of all trolleys represent the performance degradation law that generally exists under different service life and different working conditions in the whole vehicle team, and represent the overall weight change law, the influence degree of each sample is obtained by combining the overall characteristics of all influence factors of all trolleys.

[0073] As an example, the least square method is used for straight line fitting to extract the slope of the fitting straight line; in order to prevent the slope of the fitting straight line from being negative, the slope is mapped as an independent variable through the exp(x) function, and the mapping result is used as the influence factor of the sub-sequence; the average value is used to represent the overall characteristics of the influence factor, and in order to enable the influence degree to correct the training error in the subsequent process, the product of the average value of all influence factors and the influence factor of the sub-sequence to which each sample belongs is added to the sum value after the constant 1, which is used as the influence degree of each sample. The influence degree is a specific data value.

[0074] It should be noted that when the length of the sub-sequence is 1, the influence degree of the corresponding sample is 1; the average value, the median and the mode of all influence factors can also be weighted and fused to represent the overall characteristics of the influence factor; the least square method, the method of obtaining the slope and the exponential function exp(x) with the natural constant e as the base number are all known technologies and will not be described in detail.

[0075] In addition, considering that if the trolley is used multiple times, the samples whose performance is affected occupy the majority, and the time sequence first sample of the trolley is not always the first use of the trolley, the result of "clustering based on electro-hydraulic driving parameters" has a certain deviation, resulting in that the change trend of the initial training weight is not obvious enough, and the important coefficient represents the importance of the sample, so the initial training weight is also corrected by combining the training error of the current training and the important coefficient to obtain the corrected training weight of each sample in the next training process until the model converges.

[0076] Preferably, in one embodiment of the present application, please refer to Figure 3 Fig. 2 shows a flowchart of a method for obtaining the modified training weight of the next training process according to one embodiment of the present application, which specifically comprises the following steps:

[0077] Step S311: During the first training, the initial training weight is taken as the modified training weight; and the predicted parameter of the electro-hydraulic drive of each sample is obtained based on the modified training weight of the current training.

[0078] In order to facilitate the description and construction of the cycle, the initial training weight is taken as the modified training weight during the first training, so that there is a corresponding modified training weight for each training, and thus the state vector of each sample is inputted based on the modified training weight of the current training, and the predicted parameter of the electro-hydraulic drive of each sample is obtained by using the Transformer network.

[0079] Step S312: The predicted parameter is modified according to the influence degree and the importance coefficient, and the training error of each sample of the current training is obtained according to the difference between the modified predicted parameter and the electro-hydraulic drive parameter of the sample.

[0080] It is considered that the greater the influence degree, the greater the influence of the performance degradation of the trolley on the electro-hydraulic drive parameter, resulting in that the predicted electro-hydraulic drive parameter is too small, and the greater the importance coefficient of the sample, the worse the distribution concentration of the electro-hydraulic drive parameter, and the greater the influence of the performance on the electro-hydraulic drive parameter of the sample, and the greater the modification range of the predicted parameter, so the predicted parameter is modified according to the influence degree and the importance coefficient, and thus the training error of each sample of the current training is obtained according to the modified predicted parameter.

[0081] As an example, for each sample, the product of the influence degree and the importance coefficient is linearly normalized in the corresponding data dimension, the sum of the normalized result and the constant 1 is taken as the modification factor, the product of the modification factor and the predicted parameter is taken as the modified predicted parameter, the absolute value of the difference between the modified predicted parameter and the electro-hydraulic drive parameter of the sample is taken as the numerator, the electro-hydraulic drive parameter of the sample is taken as the denominator, and the fractional ratio after linear normalization in the corresponding data dimension is taken as the training error of each sample of the current training.

[0082] Among them, the influence degree and the importance coefficient are fused by multiplication, and the constant 1 is used to ensure that the modification factor is greater than 1; the absolute difference between the modified predicted parameter and the electro-hydraulic drive parameter of the sample is represented by the absolute value of the difference, and the relative difference is represented by the fractional ratio, so as to obtain the training error.

[0083] It should be noted that, in one embodiment of the present application, the data dimension corresponding to the linear normalization is the data dimension formed by the corresponding parameters of all samples.

[0084] Step S313: fusing the training error and the initial training weight to obtain the modified training weight of each sample in the next training process.

[0085] When the training error is larger, it indicates that more attention should be paid to the sample in the next training, so the training weight of the sample in the next training process is increased, and therefore the training error and the initial training weight are fused.

[0086] As an example, since the training error is normalized, the larger the training error, the more attention should be paid in the next training, and the larger the modified training weight. In order to ensure the correction effect of the training error, the sum of the training error and a constant 1 is taken as the correction coefficient, and the product of the correction coefficient of each sample and the initial training weight is taken as the modified training weight of each sample in the next training process.

[0087] Preferably, in an embodiment of the present application, the mean square error is taken as the error function of training, and when the error reaches the minimum, it is determined that the model converges, and the training of the model is stopped, and thus the trained neural network model is obtained.

[0088] Step S4: inputting the state vector of the current trolley into the trained model to obtain the electro-hydraulic driving parameter.

[0089] After the model is trained through steps S1-S3, the state vector of the current trolley can be input into the trained model to obtain the electro-hydraulic driving parameter, thereby realizing the control of the material distribution trolley.

[0090] Specifically, the working condition data of the trolley are converted into a state vector, which is input into the trained model to output the electro-hydraulic driving parameter, so as to provide more accurate and optimal trolley electro-hydraulic driving parameter and ensure energy saving effect.

[0091] An embodiment of the present application further provides an energy-saving electro-hydraulic driving material distribution trolley, which comprises a trolley body containing a hydraulic integrated module, a memory, a processor and a computer program, wherein the memory is used for storing the corresponding computer program, the processor is used for running the corresponding computer program, and the computer program can realize the hydraulic integrated module control method of the energy-saving electro-hydraulic driving material distribution trolley described in steps S1-S4 when running in the processor.

[0092] In summary, in view of the technical problem that the historical data is affected by the running state of the trolley, resulting in insufficient control accuracy through machine learning and affecting the energy-saving effect, the application provides an energy-saving electro-hydraulic drive distribution trolley and a hydraulic integrated module control method. The application first acquires historical distribution data samples of the trolley; further analyzes the distribution concentration of electro-hydraulic drive parameters in clustering based on state vectors, and acquires important coefficients; further analyzes the matching degree of samples in two types of clusters of each type of state vector in the clustering of electro-hydraulic drive parameters with limited state vector categories and unrestricted, and acquires important correction coefficients; further acquires the influence degree according to the change trend of the initial training weight; further corrects the initial training weight to obtain the correction training weight of the next training according to the training error of the current training, the influence degree and the important coefficient of the sample, until the model converges, solves the problem that the historical data is affected by the running state of the trolley, resulting in unsatisfactory machine learning results and inaccurate prediction control, so that more accurate and optimal electro-hydraulic drive parameters of the trolley can be provided in the prediction process, and the energy-saving effect is guaranteed.

[0093] It should be noted that the above-mentioned sequence of the embodiments of the application is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are also possible or may be advantageous.

[0094] Each embodiment in the specification is described in a progressive manner, and the same or similar parts between each embodiment can be referred to each other. Each embodiment mainly describes the difference from other embodiments.

Claims

1. A hydraulic integrated module control method for an energy-saving electro-hydraulic driven material distribution trolley, characterized in that, The method includes: Obtain historical material distribution data samples of the trolley, including state vectors and electro-hydraulic drive parameters; In the clustering of samples based on the state vectors, the importance coefficient of each sample is obtained according to the distribution concentration of the electro-hydraulic drive parameters within each cluster; in the samples of each class of state vectors, clustering is performed based on the electro-hydraulic drive parameters to obtain a first class of clusters; all samples are clustered based on the electro-hydraulic drive parameters to obtain a second class of clusters; the matching degree of the samples in the first class of clusters and the second class of clusters corresponding to the samples of each class of state vectors is analyzed to obtain the importance correction coefficient of each sample; the importance coefficient and the importance correction coefficient are fused to obtain the initial training weights; Based on the changing trend of the initial training weights corresponding to the same state vector of the same vehicle, the influence degree of each sample is obtained; based on the training error of the current training, combined with the influence degree of the sample and the importance coefficient, the initial training weights are corrected to obtain the corrected training weights of each sample in the next training process, until the model converges. The current state vector of the vehicle is input into the trained model to obtain the electro-hydraulic drive parameters; The method for obtaining the importance coefficient includes: for each cluster: sorting by the number of each type of electro-hydraulic driving parameter, and obtaining the importance coefficient of the samples in each cluster based on the difference in the proportion of samples of the two types of electro-hydraulic driving parameters with the most and second most combined samples; The method for obtaining the important correction coefficient includes: in the samples of each class of state vectors: performing cluster matching based on the proportion of common samples in the first class cluster and the second class cluster; and obtaining the important correction coefficient of each sample based on the similarity in quantity between the matched second class cluster and the first class cluster, combined with the proportion of common samples at the time of matching.

2. The hydraulic integrated module control method for an energy-saving electro-hydraulic driven material distribution trolley according to claim 1, characterized in that, The methods for obtaining the degree of influence include: The initial training weights of all samples of the same car are sorted in chronological order to construct a weight sequence. The initial training weights corresponding to the same state vector of each class are extracted and the order in the weight sequence is retained to construct a subsequence. Extract the slope of the fitted line for each subsequence to obtain the influence factor of each subsequence; based on the overall characteristics of all influence factors of all cars, and combined with the influence factor of the subsequence to which each sample belongs, obtain the influence degree of each sample.

3. The hydraulic integrated module control method for an energy-saving electro-hydraulic driven material distribution trolley according to claim 1, characterized in that, The method for obtaining the corrected training weights for each sample in the next training process includes: During the initial training, the initial training weights are used as the corrected training weights; the electrohydraulic drive prediction parameters for each sample are obtained based on the corrected training weights currently being trained. The prediction parameters are corrected based on the degree of influence and the importance coefficient, and the training error of each sample in the current training is obtained based on the difference between the corrected prediction parameters and the electro-hydraulic drive parameters of the sample. By combining the training error and the initial training weights, the corrected training weights for each sample in the next training process are obtained.

4. The hydraulic integrated module control method for an energy-saving electro-hydraulic driven material distribution trolley according to claim 1, characterized in that, The methods for determining model convergence include: The mean squared error is used as the training error function. When the error reaches its minimum, the model is considered to have converged, and training of the model is stopped.

5. The hydraulic integrated module control method for an energy-saving electro-hydraulic driven material distribution trolley according to claim 1, characterized in that, The method for obtaining the initial training weights includes: The product of the importance coefficient of the sample and the importance correction coefficient is used as the initial training weight.

6. The hydraulic integrated module control method for an energy-saving electro-hydraulic driven material distribution trolley according to claim 1, characterized in that, The method for obtaining the state vector includes: The historical material distribution data samples include operating condition data, specifically including: material weight, trolley speed, hydraulic system pressure, vibration, and load current. The LeNet feature extraction network is used to obtain the state vector of the operating condition data for each sample.

7. The hydraulic integrated module control method for an energy-saving electro-hydraulic driven material distribution trolley according to claim 1, characterized in that, The clustering was performed using the ISODATA algorithm.

8. An energy-saving electro-hydraulic driven material dispensing trolley, comprising a trolley body containing a hydraulic integrated module, and further comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the hydraulic integrated module control method for an energy-saving electro-hydraulic driven material distribution trolley as described in any one of claims 1 to 7.

Images