Hydraulic pump power self-adaptive control method and system based on load dynamic perception

Abstract

The invention discloses a hydraulic pump power self-adaptive control method and system based on load dynamic sensing, and relates to the technical field related to hydraulic pumps, and the method comprises the steps that a pressure sensor, a flow sensor and a displacement sensor are arranged in a hydraulic pump system, and long-time-sequence monitoring is carried out; carrying out interactive fusion analysis on the physical behavior characteristics of the hydraulic pump; carrying out hydraulic pump load scene identification; interactive guidance prediction is carried out, and a hydraulic pump power self-adaptive control template is matched; acquiring a real-time pressure monitoring data sequence, a real-time flow monitoring sequence and a real-time load monitoring data sequence; and carrying out adaptive correction on the hydraulic pump power adaptive control template, determining a real-time hydraulic pump power adaptive control scheme, and carrying out power adaptive control. The technical problems that in the prior art, multi-source information fusion is insufficient, the load scene recognition precision is low, and the self-adaptive control real-time performance is poor are solved, and the technical effects that the power of the hydraulic pump is accurately matched, the energy efficiency is dynamically optimized, and the control response speed is increased are achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of hydraulic pump, and particularly relates to a hydraulic pump power adaptive control method and system based on load dynamic perception. BACKGROUND

[0002] As the core power unit of the irrigation system, the power control of the hydraulic pump directly affects the water resource utilization efficiency, the energy efficiency, the stability and the reliability of the system. The traditional hydraulic pump power control adopts a fixed displacement or a pressure-flow regulation strategy based on a fixed threshold, which is difficult to cope with the complex and changeable load conditions in the irrigation operation, such as the water demand difference of different crop types, the change of pipeline resistance, the terrain undulation and the like, resulting in energy waste, uneven system power matching or response lag. In a large irrigation system, the hydraulic pump often faces a dynamically changing load scenario. For example, in different modes such as sprinkling irrigation, drip irrigation or canal water conveyance, the pressure and flow required by the hydraulic pump are significantly different. At the same time, the soil moisture regime, the climate condition and the crop growth stage further aggravate the time-varying and uncertainty of the load. The existing control method lacks real-time perception and dynamic analysis capability of the load state, and only feedback control is performed through local parameters, which is difficult to realize precise power adaptation, resulting in excessive energy consumption of the system in low load and insufficient response in high load. Moreover, the existing control method fails to fully exploit the interactive features between multi-source data and the coupling relationship of the load scenario, which limits the adaptability and robustness of the control strategy.

[0003] Therefore, in the related art, there are technical problems of insufficient multi-source information fusion, low load scenario recognition accuracy and poor real-time adaptability of the control. SUMMARY

[0004] The hydraulic pump power adaptive control method and system based on load dynamic perception provided in the present application solve the technical problems of insufficient multi-source information fusion, low load scenario recognition accuracy and poor real-time adaptability of the control in the prior art, and achieve the technical effects of precise matching of the hydraulic pump power, dynamic optimization of the energy efficiency and improvement of the control response speed.

[0005] The application provides a hydraulic pump power adaptive control method based on load dynamic perception, which comprises the following steps: arranging pressure sensors, flow sensors and displacement sensors in a hydraulic pump system, and using the pressure sensors, flow sensors and displacement sensors to perform long-time sequence monitoring on the hydraulic pump system to determine pressure monitoring data sequences, flow monitoring data sequences and load monitoring data sequences; performing hydraulic pump physical behavior characteristic interactive fusion analysis based on the pressure monitoring data sequences and the flow monitoring data sequences to determine hydraulic pump physical behavior fusion characteristics; performing hydraulic pump load scene identification based on the load monitoring data sequences to determine hydraulic pump load scene characteristics; performing interactive guided prediction on the hydraulic pump physical behavior fusion characteristics based on the hydraulic pump load scene characteristics, and matching a hydraulic pump power adaptive control template according to a prediction result; using the pressure sensors, flow sensors and displacement sensors to obtain real-time pressure monitoring data sequences, real-time flow monitoring sequences and real-time load monitoring data sequences of the hydraulic pump system; performing adaptive correction on the hydraulic pump power adaptive control template according to the real-time pressure monitoring data sequences, real-time flow monitoring sequences and real-time load monitoring data sequences, determining a real-time hydraulic pump power adaptive control scheme, and performing power adaptive control on the hydraulic pump system according to the real-time hydraulic pump power adaptive control scheme.

[0006] In a possible implementation, the hydraulic pump power adaptive control method based on load dynamic perception further performs the following processing: a physical behavior characteristic analyzer is pre-constructed; the physical behavior characteristic analyzer is called to perform characteristic analysis on the pressure monitoring data sequences and the flow monitoring data sequences respectively to determine pressure monitoring trend characteristics and flow monitoring trend characteristics; and the pressure monitoring trend characteristics and the flow monitoring trend characteristics are subjected to characteristic interactive fusion analysis to determine the hydraulic pump physical behavior fusion characteristics.

[0007] In a possible implementation, the hydraulic pump power adaptive control method based on load dynamic perception further performs the following processing: the physical behavior characteristic analyzer is obtained by supervising training of a framework based on a feedforward neural network through training data, and the physical behavior characteristic analyzer comprises an input layer, a convolution layer and an output layer.

[0008] In a possible implementation, the load-dynamic-aware hydraulic pump power adaptive control method further performs the following processing: calling an interaction probability analyzer to perform interaction probability analysis on the pressure monitoring trend feature and the flow monitoring trend feature to obtain an interaction probability set; constructing a feature interaction fusion matrix based on the interaction probability set, and performing feature interaction fusion on the pressure monitoring trend feature and the flow monitoring trend feature respectively by using the feature interaction fusion matrix to obtain a pressure monitoring fusion trend feature and a flow monitoring fusion trend feature; and taking the pressure monitoring fusion trend feature and the flow monitoring fusion trend feature as the hydraulic pump physical behavior fusion feature.

[0009] In a possible implementation, the load-dynamic-aware hydraulic pump power adaptive control method further performs the following processing: constructing a load scenario library, where the load scenario library includes a plurality of historical load scenario features and a plurality of historical load monitoring data record templates; matching the load monitoring data sequence with the plurality of historical load monitoring data record templates to obtain a matching historical load monitoring data record template, and taking a corresponding historical load scenario feature as the hydraulic pump load scenario feature.

[0010] In a possible implementation, the load-dynamic-aware hydraulic pump power adaptive control method further performs the following processing: obtaining a historical load monitoring data record set of a hydraulic pump system; performing same-class division on the historical load monitoring data record set to obtain a plurality of historical load monitoring data record sets; performing scene feature set extraction on the plurality of historical load monitoring data record sets to obtain a plurality of historical load scenario features; calculating a mean value of the plurality of historical load monitoring data record sets to construct a plurality of historical load monitoring data record templates; and mapping and associating the plurality of historical load scenario features with the plurality of historical load monitoring data record templates to construct a load scenario library.

[0011] In a possible implementation, the load-dynamic-aware hydraulic pump power adaptive control method further performs the following processing: taking the hydraulic pump load scenario feature as a constraint, performing interaction-guided prediction on the hydraulic pump physical behavior fusion feature by using a physical behavior predictor to obtain a prediction result; performing hydraulic power adaptive control scheme matching based on the prediction result to obtain a matching hydraulic power adaptive control scheme set; and performing screening on the matching hydraulic power adaptive control scheme set to determine a hydraulic power adaptive control template.

[0012] In a possible implementation, the hydraulic pump power adaptive control method based on dynamic load sensing further performs the following processing: calculating a mean value of the adaptive control scheme set of the matching hydraulic power, taking the mean value as a screening center, and performing iterative screening on the adaptive control scheme set of the matching hydraulic power based on the screening center to determine the hydraulic power adaptive control template.

[0013] In a possible implementation, the hydraulic pump power adaptive control method based on dynamic load sensing further performs the following processing: obtaining a preset feedback window; identifying power fluctuation of the hydraulic pump system in the preset feedback window, and obtaining a warning instruction if the power fluctuation identification result exceeds a preset threshold.

[0014] The application further provides a hydraulic pump power adaptive control system based on dynamic load sensing, which comprises: a monitoring data sequence determination module, configured to arrange pressure sensors, flow sensors, and displacement sensors in a hydraulic pump system, and perform long-time sequence monitoring on the hydraulic pump system by using the pressure sensors, flow sensors, and displacement sensors to determine pressure monitoring data sequences, flow monitoring data sequences, and load monitoring data sequences; a physical feature fusion analysis module, configured to perform interactive fusion analysis on physical behavior features of the hydraulic pump based on the pressure monitoring data sequences and the flow monitoring data sequences to determine physical behavior fusion features of the hydraulic pump; a load scene identification module, configured to perform load scene identification of the hydraulic pump based on the load monitoring data sequences to determine load scene features of the hydraulic pump; an interactive guidance prediction module, configured to perform interactive guidance prediction on the physical behavior fusion features of the hydraulic pump based on the load scene features of the hydraulic pump, and match a hydraulic pump power adaptive control template according to a prediction result; a real-time monitoring data sequence acquisition module, configured to acquire real-time pressure monitoring data sequences, real-time flow monitoring sequences, and real-time load monitoring data sequences of the hydraulic pump system by using the pressure sensors, flow sensors, and displacement sensors; and a power adaptive control module, configured to perform adaptive correction on the hydraulic pump power adaptive control template according to the real-time pressure monitoring data sequences, real-time flow monitoring sequences, and real-time load monitoring data sequences, determine a real-time hydraulic pump power adaptive control scheme, and perform power adaptive control on the hydraulic pump system according to the real-time hydraulic pump power adaptive control scheme.

[0015] The hydraulic pump power adaptive control method and system based on load dynamic perception provided in the application are used to arrange pressure sensors, flow sensors and displacement sensors in the hydraulic pump system and perform long-time sequence monitoring, perform interactive fusion analysis of the physical behavior characteristics of the hydraulic pump, perform load scene identification of the hydraulic pump, perform interactive guided prediction and match the hydraulic pump power adaptive control template, obtain real-time pressure monitoring data sequence, real-time flow monitoring sequence and real-time load monitoring data sequence, perform adaptive correction on the hydraulic pump power adaptive control template, determine the real-time hydraulic pump power adaptive control scheme and perform power adaptive control. The technical problems of insufficient multi-source information fusion, low load scene identification accuracy and poor real-time adaptive control in the prior art are solved, and the technical effects of accurately matching the hydraulic pump power, dynamically optimizing the energy efficiency and improving the control response speed are achieved. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings of the embodiments of the present application will be briefly introduced below. The flowchart is used to illustrate the operations performed by the system according to the embodiments of the present application in the present application. It should be understood that the foregoing or the following operations are not necessarily performed in sequence. On the contrary, according to the needs, various steps can be processed in reverse order or at the same time. At the same time, other operations can be added to these processes, or one or more steps of operation can be removed from these processes.

[0017] Figure 1 The hydraulic pump power adaptive control method based on load dynamic perception provided in the embodiments of the present application is shown in the flowchart.

[0018] Figure 2 The hydraulic pump power adaptive control system structure based on load dynamic perception provided in the embodiments of the present application is shown in the structural diagram.

[0019] The reference signs are explained as follows: monitoring data sequence determination module 10, physical characteristic fusion analysis module 20, load scene identification module 30, interactive guided prediction module 40, real-time monitoring data sequence acquisition module 50, power adaptive control module 60. DETAILED DESCRIPTION

[0020] The above description is only a summary of the technical solutions of the present application. In order to more clearly understand the technical means of the present application, the embodiments of the present application can be implemented according to the content of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application are described as follows.

[0021] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings, and the described embodiments should not be regarded as limitations of the present application. All other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application.

[0022] In the following description, "some embodiments" are involved, which describe a subset of all possible embodiments, but it can be understood that "some embodiments" can be the same subset or different subsets of all possible embodiments, and can be combined with each other without conflict. The term "first\second" involved only distinguishes similar objects, and does not represent a specific order of the objects. The terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or server including a series of steps or units does not have to be limited to those steps or units clearly listed, but can include other steps or modules that are not clearly listed or inherent to these processes, methods, products or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as understood by those skilled in the art to which the present application belongs. The terms used herein are only for the purpose of describing the embodiments of the present application.

[0023] The embodiments of the present application provide a hydraulic pump power adaptive control method based on load dynamic perception, as shown in Figure 1 The method comprises the following steps: Step S100, pressure sensors, flow sensors and displacement sensors are arranged in a hydraulic pump system, and the hydraulic pump system is monitored by using the pressure sensors, flow sensors and displacement sensors for a long time sequence to determine pressure monitoring data sequence, flow monitoring data sequence and load monitoring data sequence.

[0024] Preferably, by arranging pressure sensors, flow sensors and displacement sensors in the hydraulic pump system and long-time sequence monitoring of the hydraulic pump operating state, pressure monitoring data sequences, flow monitoring data sequences and load monitoring data sequences are obtained. Specifically, the pressure sensor is used to monitor the liquid pressure at the output end of the hydraulic pump, such as the water pressure in the pipeline, reflecting the system load size, with high pressure and large load, and low pressure and small load; the flow sensor is used to measure the liquid flow output by the hydraulic pump, the liquid volume per unit time, reflecting the system demand such as nozzle opening and closing and pipeline leakage; the displacement sensor is used to monitor the displacement or motion state of the hydraulic actuator, indirectly reflecting the mechanical characteristics of the load, such as the moving speed of the spray arm and the opening degree of the valve; long-time sequence monitoring refers to continuous data collection by the sensor at a fixed sampling frequency, forming time-continuous data sequences, including pressure monitoring data sequences, flow monitoring data sequences and load monitoring data sequences, and the load monitoring data sequences are load states calculated by displacement sensors and other load-related parameters, such as light load, light load and heavy load.

[0025] Step S200, based on the pressure monitoring data sequence and the flow monitoring data sequence, the physical behavior characteristics of the hydraulic pump are interactively analyzed and fused to determine the physical behavior fusion characteristics of the hydraulic pump.

[0026] Step S200 further comprises the following steps: S210, pre-constructing a physical behavior characteristic analyzer; S220, calling the physical behavior characteristic analyzer to analyze the pressure monitoring data sequence and the flow monitoring data sequence respectively to determine the pressure monitoring trend characteristics and the flow monitoring trend characteristics; S230, analyzing and fusing the pressure monitoring trend characteristics and the flow monitoring trend characteristics to determine the physical behavior fusion characteristics of the hydraulic pump.

[0027] Preferably, the physical behavior characteristic analyzer is pre-constructed based on a feedforward neural network, which is used to extract key features of the hydraulic pump operation from sensor data, i.e. to analyze the time sequence variation law of pressure and flow data and identify the physical state of the hydraulic pump, such as normal operation, efficiency decline and potential failure; the physical behavior characteristic analyzer is called to analyze the pressure monitoring data sequence to extract pressure monitoring trend characteristics such as pressure fluctuation amplitude, rising / falling slope and steady-state value, for example, continuous pressure rise may indicate nozzle blockage; the physical behavior characteristic analyzer is called to analyze the flow monitoring data sequence to extract flow monitoring trend characteristics such as flow stability, instantaneous change rate and phase difference with pressure, for example, sudden flow drop accompanied by pressure rise may indicate pipeline leakage; finally, the pressure monitoring trend characteristics and the flow monitoring trend characteristics are analyzed and fused interactively, i.e. the correlation strength such as covariance and mutual information between the pressure and flow characteristics is calculated, the interaction probability is determined and converted into a weight matrix, the pressure monitoring trend characteristics and the flow monitoring trend characteristics are weighted and fused, and the physical behavior fusion characteristics of the hydraulic pump are obtained.

[0028] Further, step S210 further includes that the physical behavior feature analyzer is obtained by supervising training of a framework constructed based on a feedforward neural network, and the physical behavior feature analyzer includes an input layer, a convolution layer and an output layer.

[0029] Preferably, the physical behavior feature analyzer is a framework constructed based on a feedforward neural network and trained by supervised learning, and is used to extract high-order features from sensor data of the hydraulic pump, that is, the historical data set containing pressure and flow sequence and corresponding expert labeled features is used for training, including forward propagation, input data is calculated by network to obtain predicted features, and back propagation adjusts weights by gradient descent to minimize the error between predicted features and real features, and after training, pressure / flow trend features can be automatically extracted from new data. The core structure includes an input layer, a convolution layer and an output layer, wherein the input layer is used to receive long-time pressure monitoring data sequence and flow monitoring data sequence and perform preprocessing, including normalization processing and sliding window segmentation, and the long sequence is cut into fixed-length subsequences; the convolution layer is used to extract local space-time features of sensor data, such as pressure surge and flow periodic fluctuation, and the convolution kernel slides along the time axis to capture short-term patterns and generate multiple feature representations of different behavior patterns; and the output layer integrates the local features extracted by the convolution layer into global trend features, including compressing feature dimensions, generating pressure monitoring trend features and flow monitoring trend features.

[0030] Further, step S210 further includes step S211 of calling an interaction probability analyzer to perform interaction probability analysis on the pressure monitoring trend features and the flow monitoring trend features to obtain an interaction probability set; step S212 of constructing a feature interaction fusion matrix based on the interaction probability set, and performing feature interaction fusion on the pressure monitoring trend features and the flow monitoring trend features by using the feature interaction fusion matrix respectively to obtain pressure monitoring fusion trend features and flow monitoring fusion trend features; and step S213 of taking the pressure monitoring fusion trend features and the flow monitoring fusion trend features as the physical behavior fusion features of the hydraulic pump.

[0031] Preferably, the interaction probability analyzer is called to analyze the interaction probability of the pressure monitoring trend feature and the flow monitoring trend feature, wherein the interaction probability analyzer is essentially a similarity calculator, which is embedded with a cosine similarity function to calculate the cosine similarity of the pressure monitoring trend feature and the flow monitoring trend feature, representing the direction consistency of the two feature vectors, with a value range of [-1, 1], and the closer to 1, the higher the similarity, that is, the higher the correlation between the two features, and the higher the corresponding interaction probability, thereby obtaining an interaction probability set containing multiple similarity scores; the interaction probability set is normalized, and an empty matrix of MxN is constructed, wherein N is the pressure feature dimension and M is the flow feature dimension, and then the similarity of the normalized pressure feature and the flow feature is filled into the empty matrix as a matrix element to obtain a feature interaction fusion matrix; finally, the feature interaction fusion matrix is used to perform feature interaction fusion on the pressure monitoring trend feature and the flow monitoring trend feature, that is, the convolutional neural network is used to enhance the implicit information in the original features, including weighted summation of the pressure feature vector and the row vector of the matrix, and similarly, the column vector of the matrix is used to weight the flow feature, and the pressure monitoring fusion trend feature and the flow monitoring fusion trend feature are output; finally, the pressure monitoring fusion trend feature and the flow monitoring fusion trend feature are spliced and fused to obtain a hydraulic pump physical behavior fusion feature representing the comprehensive state of the hydraulic pump, thereby quantifying the dynamic relationship between the pressure and flow of the hydraulic pump, and ensuring the reliability and energy efficiency of the hydraulic system.

[0032] Step S300, based on the load monitoring data sequence, a hydraulic pump load scene recognition is performed to determine a hydraulic pump load scene feature.

[0033] Step S300 further includes step S310 of constructing a load scene library, wherein the load scene library includes a plurality of historical load scene features and a plurality of historical load monitoring data record templates; step S320, the load monitoring data sequence is matched with the plurality of historical load monitoring data record templates to obtain a matched historical load monitoring data record template, and the corresponding historical load scene feature is taken as the hydraulic pump load scene feature.

[0034] Preferably, the load scene library is constructed with multiple historical load scene features and multiple historical load monitoring data record templates, wherein the historical load scene features refer to typical working condition feature vectors extracted from the historical operation data of the hydraulic pump, such as "light load uniform speed", "heavy load impact", etc., and the historical load monitoring data record template is a standardized data pattern corresponding to the scene feature, such as the time sequence law of pressure, flow rate, displacement; then the load monitoring data sequence is matched with the multiple historical load monitoring data record templates, for example, the similarity of the load monitoring data sequence and each historical load monitoring data record template is calculated using dynamic time warping or Euclidean distance, the historical load monitoring data record template with high similarity is output as the matched historical load monitoring data record template, and the corresponding historical load scene feature is taken as the load scene feature of the hydraulic pump under the current working condition.

[0035] Further, step S310 further includes step S311 of acquiring a historical load monitoring data record set of the hydraulic pump system; step S312 of performing same-class division on the historical load monitoring data record set to obtain multiple historical load monitoring data record sets; step S313 of performing scene feature set extraction on the multiple historical load monitoring data record sets to obtain multiple historical load scene features; step S314 of calculating the mean of the multiple historical load monitoring data record sets to construct multiple historical load monitoring data record templates; and step S315 of mapping and associating the multiple historical load scene features with the multiple historical load monitoring data record templates to construct a load scene library.

[0036] Preferably, the historical load monitoring data record set of the hydraulic pump system is acquired, that is, the sensor data of the hydraulic pump under different working conditions accumulated for a long time, including pressure sensor data, flow rate sensor data, displacement sensor data, and time stamp, environmental parameters, etc., then the historical load monitoring data record set is divided into same-class sets through K-means clustering, such as light load, heavy load, transient impact, etc., to obtain multiple historical load monitoring data record sets; then scene feature set extraction is performed on the multiple historical load monitoring data record sets, including extracting statistical features such as mean and variance of various data, behavior features such as pressure fluctuation range, flow rate mean, displacement change rate, load change rate, flow rate-pressure coupling coefficient, etc., to further obtain multiple historical load scene features; the mean of the multiple historical load monitoring data record sets is calculated respectively to construct multiple historical load monitoring data record templates, and finally the multiple historical load scene features are mapped and associated with the multiple historical load monitoring data record templates to form a structured query library, that is, a load scene library, which can quickly match the templates when new data is input, call the associated scene features to realize working condition recognition, such as automatically switching to high power mode when matching the heavy load template, so as to improve the power control response speed and energy efficiency of the hydraulic system.

[0037] Step S400, based on the hydraulic pump load scene characteristics of the physical behavior fusion characteristics of the interactive guide prediction, and according to the prediction result matching hydraulic pump power adaptive control template.

[0038] Step S400 further includes step S410, using the physical behavior predictor to predict the physical behavior fusion characteristics of the hydraulic pump under the constraint of the hydraulic pump load scene characteristics, and obtaining the prediction result; step S420, based on the prediction result, matching the hydraulic power adaptive control scheme, obtaining the matching hydraulic power adaptive control scheme set; step S430, screening the matching hydraulic power adaptive control scheme set, and determining the hydraulic power adaptive control template.

[0039] Preferably, the hydraulic pump load scene characteristics are used as constraints to limit the prediction range, ensure that the prediction result conforms to the physical law of the current operating condition, and use the physical behavior predictor to predict the physical behavior fusion characteristics of the hydraulic pump, wherein the physical behavior predictor is a prediction model based on neural network model or LSTM time series model, used to analyze the evolution trend of the physical behavior fusion characteristics under the current load scene, and can obtain short-term state prediction and key indicator prediction, which is used as the prediction result, and the possible state of pressure and flow under the current load scene. According to the prediction result, the hydraulic power adaptive control scheme is matched, that is, the control strategy meeting the conditions is selected from the preset scheme library, a plurality of matching hydraulic power adaptive control schemes are obtained, and a matching hydraulic power adaptive control scheme set is formed; finally, based on the priority of safety> energy efficiency> response speed, the matching hydraulic power adaptive control scheme set is screened, the optimal solution in the matching hydraulic power adaptive control scheme set is calculated, and the hydraulic power adaptive control template is finally determined.

[0040] Further, step S430 further includes calculating the mean value of the matching hydraulic power adaptive control scheme set, taking it as the screening center, and based on the screening center, iteratively screening the matching hydraulic power adaptive control scheme set to determine the hydraulic power adaptive control template.

[0041] Preferably, the matching hydraulic power adaptive control scheme set is screened by a mean shift algorithm, wherein the mean shift algorithm is a clustering algorithm that drifts to the local maximum density by iteratively calculating the density gradient of data points, and finally converges to the center of data distribution. The control scheme set is regarded as a point in a multi-dimensional space, such as dimensions including power adjustment amplitude, pressure limiting value, etc. The most dense area is found by mean shift, representing the optimal solution recognized by the majority of schemes. Specifically, the mean of the matching hydraulic power adaptive control scheme set is calculated and taken as the iteration starting point (i.e. screening center), and the matching hydraulic power adaptive control scheme set is iteratively screened based on the screening center, including setting a radius (such as an Euclidean distance threshold) with the current center point as the center to circumscribe the schemes in the neighborhood; calculating the mean of all schemes in the neighborhood as the new center point; repeating the drift process until the center point changes less than the threshold, then stopping iteration, and finally outputting the converged point as the hydraulic power adaptive control template.

[0042] Step S500, using the pressure sensor, flow sensor and displacement sensor, acquiring the real-time pressure monitoring data sequence, real-time flow monitoring sequence and real-time load monitoring data sequence of the hydraulic pump system.

[0043] Preferably, the pressure sensor is used to detect the liquid pressure at the outlet of the hydraulic pump in real time, reflecting the load intensity of the hydraulic pump system. A continuous rise in pressure may indicate a blockage. The flow sensor is used to monitor the volume flow of the liquid in the hydraulic circuit, representing the functional demand of the hydraulic pump system. A decrease in flow may indicate a leak. The displacement sensor is used to track the displacement of the actuator (such as the hydraulic cylinder piston), indirectly reflecting the change in mechanical load. Accelerated displacement may indicate a reduction in load. The three types of sensor data are synchronously collected at the same sampling frequency to form time-aligned monitoring data sequences, including real-time pressure monitoring data sequence, real-time flow monitoring sequence and real-time load monitoring data sequence.

[0044] Step S600, according to the real-time pressure monitoring data sequence, real-time flow monitoring sequence and real-time load monitoring data sequence, adaptively correcting the hydraulic pump power adaptive control template, determining a real-time hydraulic pump power adaptive control scheme, and performing power adaptive control on the hydraulic pump system according to the real-time hydraulic pump power adaptive control scheme.

[0045] Preferably, the pre-generated hydraulic pump power adaptive control template is dynamically fine-tuned through the real-time sensor data stream so that the hydraulic pump output power always accurately matches the current load demand. Specifically, the real-time pressure monitoring data sequence, the real-time flow monitoring sequence and the real-time load monitoring data sequence are compared with the expected hydraulic pump power adaptive control template to detect the deviation value of the hydraulic pump system load parameter, and then the hydraulic pump power adaptive control template is adaptively corrected, that is, the hydraulic pump power adaptive control template parameters are dynamically corrected through the PID algorithm or fuzzy logic, such as adjusting the power increase range from 8% to 10%, and then determining the real-time hydraulic pump power adaptive control scheme. Finally, the hydraulic pump system is power-adaptively controlled according to the real-time hydraulic pump power adaptive control scheme, which may include a PLC or embedded controller receiving the real-time hydraulic pump power adaptive control scheme and adjusting the frequency converter (motor speed) and proportional valve (flow) of the hydraulic pump system to achieve power adaptive control of the hydraulic pump system.

[0046] Furthermore, step S600 also includes step S610, obtaining a preset feedback window; step S620, identifying power fluctuations of the hydraulic pump system in the preset feedback window, and obtaining a warning instruction if the power fluctuation identification result exceeds a preset threshold.

[0047] Preferably, a time window of fixed length is set as a preset feedback window for periodically evaluating the power stability of the hydraulic pump, and then the power fluctuation of the hydraulic pump system is identified in the preset feedback window, that is, the standard deviation of the power sequence in the window is calculated to obtain the fluctuation amplitude of the power in the window. The preset threshold is the power fluctuation range set according to historical operating data or equipment safety limit. If the power fluctuation identification result exceeds the preset threshold, it is determined to be an abnormal power fluctuation and an early warning instruction is triggered. The early warning type may include a first-level early warning, where the power fluctuation exceeds the threshold but is not dangerous, a log is recorded, and the operator is prompted to check; a second-level early warning, where the power fluctuation continues to deteriorate, and the power is automatically reduced or shut down; for example, if the power fluctuation of the hydraulic pump system is detected to exceed the limit within 10 seconds, it will automatically switch to safety mode and alarm.

[0048] In the above, refer to Figure 1 The hydraulic pump power adaptive control method based on load dynamic perception according to an embodiment of the present invention is described in detail. Figure 2 A hydraulic pump power adaptive control system based on load dynamic perception according to an embodiment of the present invention is described.

[0049] The hydraulic pump power adaptive control system based on dynamic load perception according to the embodiment of the present invention is used to solve the technical problems existing in the prior art such as insufficient multi-source information fusion, low load scene recognition accuracy, and poor real-time performance of adaptive control, and achieves the technical effects of accurately matching hydraulic pump power, dynamically optimizing energy efficiency, and improving control response speed.Figure 2 As shown, the hydraulic pump power adaptive control system based on load dynamic perception comprises a monitoring data sequence determination module 10, a physical feature fusion analysis module 20, a load scene identification module 30, an interaction guide prediction module 40, a real-time monitoring data sequence acquisition module 50, and a power adaptive control module 60.

[0050] The monitoring data sequence determination module 10 is used to arrange pressure sensors, flow sensors and displacement sensors in the hydraulic pump system, and use the pressure sensors, flow sensors and displacement sensors to monitor the hydraulic pump system for a long time to determine pressure monitoring data sequences, flow monitoring data sequences and load monitoring data sequences. The physical feature fusion analysis module 20 is used to perform interaction fusion analysis of the physical behavior features of the hydraulic pump based on the pressure monitoring data sequences and the flow monitoring data sequences to determine the physical behavior fusion features of the hydraulic pump. The load scene identification module 30 is used to identify the load scene of the hydraulic pump based on the load monitoring data sequences to determine the load scene features of the hydraulic pump. The interaction guide prediction module 40 is used to perform interaction guide prediction on the physical behavior fusion features of the hydraulic pump based on the load scene features of the hydraulic pump, and match the hydraulic pump power adaptive control template according to the prediction result. The real-time monitoring data sequence acquisition module 50 is used to acquire real-time pressure monitoring data sequences, real-time flow monitoring sequences and real-time load monitoring data sequences of the hydraulic pump system by using the pressure sensors, flow sensors and displacement sensors. The power adaptive control module 60 is used to adaptively correct the hydraulic pump power adaptive control template according to the real-time pressure monitoring data sequences, real-time flow monitoring sequences and real-time load monitoring data sequences, determine a real-time hydraulic pump power adaptive control scheme, and perform power adaptive control on the hydraulic pump system according to the real-time hydraulic pump power adaptive control scheme.

[0051] In the following, the specific configuration of the physical feature fusion analysis module 20 will be described in detail. The physical feature fusion analysis module 20 further comprises a pre-constructed physical behavior feature analyzer, and the physical behavior feature analyzer is obtained by supervising training a framework based on a feedforward neural network by using training data. The physical behavior feature analyzer comprises an input layer, a convolution layer and an output layer.

[0052] In the following, the specific configuration of the physical feature fusion analysis module 20 will be described in detail. The physical feature fusion analysis module 20 further comprises a pre-constructed physical behavior feature analyzer, and the physical behavior feature analyzer is obtained by supervising training a framework based on a feedforward neural network by using training data. The physical behavior feature analyzer comprises an input layer, a convolution layer and an output layer.

[0053] In the following, the specific configuration of the physical feature fusion analysis module 20 will be described in detail. The physical feature fusion analysis module 20 further comprises: calling an interaction probability analyzer to perform interaction probability analysis on the pressure monitoring trend feature and the flow monitoring trend feature, and obtaining an interaction probability set; constructing a feature interaction fusion matrix based on the interaction probability set, and performing feature interaction fusion on the pressure monitoring trend feature and the flow monitoring trend feature respectively by using the feature interaction fusion matrix, to obtain a pressure monitoring fusion trend feature and a flow monitoring fusion trend feature; and taking the pressure monitoring fusion trend feature and the flow monitoring fusion trend feature as the hydraulic pump physical behavior fusion feature.

[0054] In the following, the specific configuration of the load scenario recognition module 30 will be described in detail. The load scenario recognition module 30 further comprises: constructing a load scenario library, wherein the load scenario library comprises a plurality of historical load scenario features and a plurality of historical load monitoring data record templates; matching the load monitoring data sequence with the plurality of historical load monitoring data record templates to obtain a matching historical load monitoring data record template, and taking the corresponding historical load scenario feature as the hydraulic pump load scenario feature.

[0055] In the following, the specific configuration of the load scenario recognition module 30 will be described in detail. The load scenario recognition module 30 further comprises: obtaining a set of historical load monitoring data records of the hydraulic pump system; performing same-class division on the set of historical load monitoring data records to obtain a plurality of sets of historical load monitoring data records; performing scene feature set extraction on the plurality of sets of historical load monitoring data records to obtain a plurality of historical load scenario features; calculating the mean of the plurality of sets of historical load monitoring data records to construct a plurality of historical load monitoring data record templates; and mapping and associating the plurality of historical load scenario features with the plurality of historical load monitoring data record templates to construct a load scenario library.

[0056] In the following, the specific configuration of the interaction guide prediction module 40 will be described in detail. The interaction guide prediction module 40 further comprises: using the hydraulic pump load scenario feature as a constraint, performing interaction guide prediction on the hydraulic pump physical behavior fusion feature by using a physical behavior predictor to obtain a prediction result; performing hydraulic power adaptive control scheme matching based on the prediction result to obtain a set of matched hydraulic power adaptive control schemes; and screening the set of matched hydraulic power adaptive control schemes to determine a hydraulic power adaptive control template.

[0057] Below, the specific configuration of the interaction guidance prediction module 40 will be described in detail. The interaction guidance prediction module 40 further comprises: calculating the matching hydraulic power adaptive control scheme set mean value, taking it as a screening center, and based on the screening center, iteratively screening the matching hydraulic power adaptive control scheme set, to determine the hydraulic power adaptive control template.

[0058] Below, the specific configuration of the power adaptive control module 60 will be described in detail. The power adaptive control module 60 further comprises: obtaining a preset feedback window; identifying power fluctuation of the hydraulic pump system in the preset feedback window, and if the power fluctuation identification result exceeds a preset threshold, obtaining a warning instruction.

[0059] The hydraulic pump power adaptive control system based on load dynamic perception provided by the embodiment of the application can execute the hydraulic pump power adaptive control method based on load dynamic perception provided by any embodiment of the application, and has the corresponding function modules and beneficial effects of executing the method.

[0060] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, however, any number of different modules can be used and run on the user terminal and / or server, and the various units and modules included are only divided according to the functional logic, but are not limited to the above division, as long as the corresponding functions can be realized; in addition, the specific names of the functional units are only for easy mutual differentiation, and do not limit the protection scope of the present application.

[0061] The above specific embodiments do not constitute a limitation on the protection scope of the present application. Those skilled in the art should understand that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for power adaptive control of a hydraulic pump based on dynamic load sensing, characterized in that, The method comprises: arranging pressure sensors, flow sensors and displacement sensors in a hydraulic pump system, and using the pressure sensors, flow sensors and displacement sensors to perform long-time sequence monitoring on the hydraulic pump system to determine pressure monitoring data sequences, flow monitoring data sequences and load monitoring data sequences; performing hydraulic pump physical behavior characteristic interactive fusion analysis based on the pressure monitoring data sequences and the flow monitoring data sequences to determine hydraulic pump physical behavior fusion characteristics; performing hydraulic pump load scene identification based on the load monitoring data sequences to determine hydraulic pump load scene characteristics; performing interactive guidance prediction on the hydraulic pump physical behavior fusion characteristics based on the hydraulic pump load scene characteristics, and matching a hydraulic pump power adaptive control template according to a prediction result; using the pressure sensors, flow sensors and displacement sensors to obtain real-time pressure monitoring data sequences, real-time flow monitoring sequences and real-time load monitoring data sequences of the hydraulic pump system; performing adaptive correction on the hydraulic pump power adaptive control template according to the real-time pressure monitoring data sequences, real-time flow monitoring sequences and real-time load monitoring data sequences, determining a real-time hydraulic pump power adaptive control scheme, and performing power adaptive control on the hydraulic pump system according to the real-time hydraulic pump power adaptive control scheme.

2. The method of claim 1, wherein, The hydraulic pump physical behavior characteristic interactive fusion analysis based on the pressure monitoring data sequences and the flow monitoring data sequences to determine hydraulic pump physical behavior fusion characteristics comprises: pre-constructing a physical behavior characteristic analyzer; calling the physical behavior characteristic analyzer to perform characteristic analysis on the pressure monitoring data sequences and the flow monitoring data sequences respectively to determine pressure monitoring trend characteristics and flow monitoring trend characteristics; performing characteristic interactive fusion analysis on the pressure monitoring trend characteristics and the flow monitoring trend characteristics to determine the hydraulic pump physical behavior fusion characteristics.

3. The method of claim 2, wherein the load-dynamic-aware hydraulic pump power adaptive control method is characterized by, The physical behavior characteristic analyzer is obtained by supervising training on a framework constructed based on a feedforward neural network through training data, and the physical behavior characteristic analyzer comprises an input layer, a convolution layer and an output layer.

4. The method of claim 2, wherein the load-dynamic-aware hydraulic pump power adaptive control method is characterized by, The characteristic interactive fusion analysis on the pressure monitoring trend characteristics and the flow monitoring trend characteristics to determine the hydraulic pump physical behavior fusion characteristics comprises: calling an interactive probability analyzer to perform interactive probability analysis on the pressure monitoring trend characteristics and the flow monitoring trend characteristics to obtain an interactive probability set; constructing a characteristic interactive fusion matrix based on the interactive probability set, and using the characteristic interactive fusion matrix to perform characteristic interactive fusion on the pressure monitoring trend characteristics and the flow monitoring trend characteristics respectively to obtain pressure monitoring fusion trend characteristics and flow monitoring fusion trend characteristics; using the pressure monitoring fusion trend characteristics and the flow monitoring fusion trend characteristics as the hydraulic pump physical behavior fusion characteristics.

5. The method for load-dynamic-aware hydraulic pump power adaptive control as claimed in claim 1, wherein, The hydraulic pump load scene identification based on the load monitoring data sequences to determine hydraulic pump load scene characteristics comprises: constructing a load scene library, wherein the load scene library comprises a plurality of historical load scene characteristics and a plurality of historical load monitoring data record templates; Matching the load monitoring data sequence with the plurality of historical load monitoring data record templates to obtain a matching historical load monitoring data record, and taking the historical load scene feature corresponding to the matching historical load monitoring data record as the hydraulic pump load scene feature.

6. The load-dynamic-aware hydraulic pump power adaptive control method of claim 5, wherein, Comprise: Obtain a set of historical load monitoring data records of a hydraulic pump system; Classify the set of historical load monitoring data records into the same category to obtain a plurality of sets of historical load monitoring data records; Extract scene features from the plurality of sets of historical load monitoring data records to obtain a plurality of historical load scene features; Calculate the mean of the plurality of sets of historical load monitoring data records to construct a plurality of historical load monitoring data record templates; Map the plurality of historical load scene features to the plurality of historical load monitoring data record templates to construct a load scene library.

7. The load-dynamic-aware hydraulic pump power adaptive control method of claim 1, wherein, Based on the hydraulic pump load scene feature, interactively guide the prediction of the hydraulic pump physical behavior fusion feature, and match the hydraulic pump power adaptive control template according to the prediction result, comprising: Using the hydraulic pump load scene feature as a constraint, interactively guide the prediction of the hydraulic pump physical behavior fusion feature using a physical behavior predictor to obtain a prediction result; Based on the prediction result, match the hydraulic power adaptive control scheme to obtain a set of matching hydraulic power adaptive control schemes; Screen the set of matching hydraulic power adaptive control schemes to determine the hydraulic power adaptive control template.

8. The load-dynamic-aware hydraulic pump power adaptive control method of claim 7, wherein, Calculate the mean of the set of matching hydraulic power adaptive control schemes, take it as the screening center, and based on the screening center, iteratively screen the set of matching hydraulic power adaptive control schemes to determine the hydraulic power adaptive control template.

9. The load-dynamic-aware hydraulic pump power adaptive control method of claim 1, wherein, Obtain a preset feedback window; In the preset feedback window, identify the power fluctuation of the hydraulic pump system, and if the power fluctuation identification result exceeds the preset threshold, obtain a warning instruction.

10. A load-dynamic-aware hydraulic pump power adaptive control system, characterized in that, The system is used to implement the load dynamic perception-based hydraulic pump power adaptive control method of any one of claims 1 to 9, and the system comprises: A monitoring data sequence determination module is used to arrange pressure sensors, flow sensors and displacement sensors in a hydraulic pump system, and use the pressure sensors, flow sensors and displacement sensors to monitor the hydraulic pump system for a long time sequence to determine pressure monitoring data sequences, flow monitoring data sequences and load monitoring data sequences. A physical feature fusion analysis module is used to interactively fuse and analyze the hydraulic pump physical behavior features based on the pressure monitoring data sequences and the flow monitoring data sequences to determine the hydraulic pump physical behavior fusion features. A load scene identification module is used to identify the hydraulic pump load scene based on the load monitoring data sequence to determine the hydraulic pump load scene feature. An interactive guided prediction module is used to interactively guide the prediction of the hydraulic pump physical behavior fusion feature based on the hydraulic pump load scene feature, and match the hydraulic pump power adaptive control template according to the prediction result. A real-time monitoring data sequence acquisition module is configured to acquire real-time pressure monitoring data sequences, real-time flow monitoring sequences and real-time load monitoring data sequences of the hydraulic pump system by using the pressure sensor, the flow sensor and the displacement sensor; A power adaptive control module is configured to adaptively correct the hydraulic pump power adaptive control template according to the real-time pressure monitoring data sequences, the real-time flow monitoring sequences and the real-time load monitoring data sequences, determine a real-time hydraulic pump power adaptive control scheme, and perform power adaptive control on the hydraulic pump system according to the real-time hydraulic pump power adaptive control scheme.

Images