Environmental protection equipment control method, system, equipment and medium

Abstract

The invention provides an environmental protection equipment control method and system, equipment and a medium, and belongs to the technical field of environmental protection equipment control. The method comprises the following steps: setting control parameters such as a working temperature threshold value, a flow regulation range and a pressure safety limit value, then collecting operation state data of each execution unit of the controlled environmental protection equipment, and carrying out feature extraction coding on the control parameters to obtain a low-dimensional feature coding vector; layered filtering processing based on feature weights is carried out on the operation state data to obtain key feature aggregation vectors, then association matching analysis is carried out on the key feature aggregation vectors and the historical data to obtain association matching coding vectors, and finally the association matching coding vectors are input into a decision maker to generate control instructions to indicate equipment operation adjustment. According to the method, through accurate data processing and correlation analysis, the control precision and adaptability of the environmental protection equipment are improved, the equipment operation efficiency and the environmental protection effect are improved, and the method has high practicability and popularization value.

Description

Technical Field

[0001] The present application belongs to the technical field of environmental protection equipment control, and more specifically, relates to an environmental protection equipment control method, system, equipment and medium. Background Art

[0002] With increasing environmental protection requirements, environmental protection equipment is increasingly being used in industrial production, environmental governance, and other fields. However, existing environmental protection equipment control methods have many shortcomings in practical applications. Traditional control methods typically use fixed control parameter settings, which are difficult to flexibly adjust to the different operating conditions and external environmental changes of environmental protection equipment. This leads to low equipment operating efficiency and the inability to fully realize its environmental performance. For example, when faced with operating temperature fluctuations, flow changes, or abnormal pressure, fixed parameter settings may cause the equipment to operate in a suboptimal state, which not only affects the treatment effect of pollutants but also increases energy consumption.

[0003] Existing technologies lack effective data processing mechanisms after collecting operational status data from environmental protection equipment execution units. This large amount of operational status data contains a significant amount of redundant information and irrelevant features. Directly analyzing this data not only increases the processing workload and computational complexity, but can also lead to inaccurate analysis results due to interference from irrelevant information. This inability to accurately reflect the equipment's true operational status can affect the generation of control instructions and the regulation of the equipment.

[0004] Traditional control methods often fail to fully leverage the value of historical data when analyzing the correlation between control parameters and operating status data. Correlation analysis without historical data makes it difficult to accurately grasp the inherent connection between control parameters and operating status. Consequently, the generated control instructions cannot accurately guide equipment operational adjustments. This results in poor adaptability to complex operating conditions and makes it difficult to achieve efficient and stable operation.

[0005] Existing control command generation methods are relatively simple, typically based on a single decision-making model or rule, and fail to comprehensively consider the impact of multiple factors on equipment operation. This simplistic generation method results in insufficient flexibility and accuracy in control commands, making it difficult to meet the control needs of environmental protection equipment in different scenarios, and thus limiting further improvements in environmental protection equipment performance. Summary of the Invention

[0006] The purpose of this application is to provide an environmental protection equipment control method, system, equipment and medium to solve the problems raised in the above background technology.

[0007] To achieve the above object, the present invention provides an environmental protection equipment control method, the method comprising: Setting control parameters, including operating temperature threshold, flow rate adjustment range, and pressure safety limit; Collecting the operating status data of each execution unit of the controlled environmental protection equipment under the constraints of the control parameters to obtain a set of the operating status data of the execution units; Performing feature extraction and encoding on the control parameter to obtain a low-dimensional feature encoding vector of the control parameter; Performing a hierarchical filtering process based on feature weights on the set of execution unit operation status data to obtain an execution unit operation status key feature aggregation vector; Performing an association matching analysis based on historical data enhancement on the key feature aggregation vector of the execution unit operation state and the low-dimensional feature coding vector of the control parameter to obtain a control parameter-operation state association matching coding vector; Based on the control parameter-operating state association matching coding vector, an environmental protection equipment control instruction is generated, specifically: the control parameter-operating state association matching coding vector is input into a control instruction generation module based on a decision maker to obtain an equipment control instruction, and the equipment control instruction is used to instruct the operation adjustment operation of the environmental protection equipment.

[0008] In one embodiment of the present application, performing a hierarchical filtering process based on feature weights on the set of execution unit running state data to obtain an execution unit running state key feature aggregation vector includes: Passing each execution unit operation status data in the set of execution unit operation status data through a filtering layer to obtain a set of execution unit operation status feature filtering coding vectors; Calculating a pseudo-reference filtered center representation vector of the execution unit running state based on a feature distribution density of a set of the execution unit running state feature filtering code vectors; Calculating a filtering adjustment direction of each execution unit operation state feature filtering code vector in the set of the execution unit operation state feature filtering code vectors relative to the execution unit operation state pseudo-reference filtering center representation vector to obtain a set of execution unit operation state filtering adjustment directions; Based on the set of execution unit operation state filtering adjustment directions, the set of execution unit operation state feature filtering encoding vectors is hierarchically filtered toward the execution unit operation state pseudo-reference filtering center representation vector to obtain the execution unit operation state key feature aggregation vector.

[0009] In one embodiment of the present application, calculating the execution unit running state pseudo-reference filtered center representation vector based on the feature distribution density of the set of execution unit running state feature filtering code vectors includes: Inputting each execution unit operation state feature filter code vector in the set of execution unit operation state feature filter code vectors into a feature distribution density measurement module to obtain a set of execution unit operation state distribution density measurement coefficients; Inputting the set of execution unit operation state distribution density measurement coefficients into a feature density screening control unit to obtain a set of execution unit operation state distribution density weight factors; Based on the set of execution unit operation state distribution density weight factors, a weighted sum of the set of execution unit operation state feature filter coding vectors is calculated to obtain the execution unit operation state pseudo-reference filter center representation vector.

[0010] In one embodiment of the present application, each execution unit operation state feature filter code vector in the set of the execution unit operation state feature filter code vectors is respectively input into a feature distribution density measurement module to obtain a set of execution unit operation state distribution density measurement coefficients, including: Normalizing the execution unit operation state feature filtering code vector to obtain a normalized execution unit operation state feature filtering code vector; The execution unit operation state distribution density measurement coefficient is obtained by calculating the sum of squares of each eigenvalue in the normalized execution unit operation state feature filtering code vector and dividing it by the feature dimension value of the normalized execution unit operation state feature filtering code vector.

[0011] In one embodiment of the present application, performing an association matching analysis based on historical data enhancement on the key feature aggregation vector of the execution unit operating state and the low-dimensional feature encoding vector of the control parameter to obtain a control parameter-operating state association matching encoding vector includes: Based on historical data, performing feature association optimization based on a matching mechanism on the key feature aggregation vector of the execution unit operation state and the low-dimensional feature encoding vector of the control parameter to obtain a historical data optimized control parameter-operation state association matching feature matrix; Optimizing the control parameter-operation state association matching feature matrix based on the historical data, performing feature adjustment optimization on the execution unit operation state key feature aggregation vector and the control parameter low-dimensional feature encoding vector to obtain an optimized execution unit operation state key feature aggregation vector and an optimized control parameter low-dimensional feature encoding vector; The bitwise product between the optimized execution unit running state key feature aggregation vector and the optimized control parameter low-dimensional feature encoding vector is calculated to obtain the control parameter-running state association matching encoding vector.

[0012] In one embodiment of the present application, based on historical data, feature association optimization is performed on the key feature aggregation vector of the execution unit operating state and the low-dimensional feature encoding vector of the control parameter based on a matching mechanism to obtain a historical data optimized control parameter-operating state association matching feature matrix, including: Inputting the key feature aggregation vector of the execution unit operation state and the low-dimensional feature encoding vector of the control parameter into a fine-grained association analysis network to obtain a control parameter-operation state association matching feature matrix; The control parameter-operating state association matching characteristic matrix is ​​input into a matching unit based on historical data to obtain the historical data optimized control parameter-operating state association matching characteristic matrix.

[0013] In one embodiment of the present application, based on the historical data optimization control parameter-operating state association matching feature matrix, feature adjustment optimization is performed on the execution unit operating state key feature aggregation vector and the control parameter low-dimensional feature encoding vector to obtain an optimized execution unit operating state key feature aggregation vector and an optimized control parameter low-dimensional feature encoding vector, including: performing a linear transformation on the execution unit operating state key feature aggregation vector to obtain a first correlation feature vector and a first adjustment feature vector, using the historical data optimized control parameter-operating state correlation matching feature matrix as a reference matrix, inputting the first correlation feature vector, the first adjustment feature vector, and the reference matrix into a fine-grained adjustment module based on a memory network to obtain the optimized execution unit operating state key feature aggregation vector; The control parameter low-dimensional feature encoding vector is linearly transformed to obtain a second associated feature vector and a second adjusted feature vector, and the historical data optimized control parameter-operating state associated matching feature matrix is ​​used as a reference matrix. The second associated feature vector, the second adjusted feature vector and the reference matrix are input into the memory network-based fine-grained adjustment module to obtain the optimized control parameter low-dimensional feature encoding vector.

[0014] The present invention also includes an environmental protection equipment control system for implementing the above-mentioned environmental protection equipment control method, the system comprising: A control parameter setting module is used to set control parameters, including an operating temperature threshold, a flow rate adjustment range, and a pressure safety limit; An operation status data collection module, configured to collect the operation status data of each execution unit of the controlled environmental protection equipment under the constraints of the control parameters to obtain a set of execution unit operation status data; A control parameter encoding module, configured to perform feature extraction and encoding on the control parameter to obtain a low-dimensional feature encoding vector of the control parameter; a feature hierarchical filtering processing module, configured to perform a hierarchical filtering process on the set of the execution unit running state data based on feature weights to obtain a key feature aggregation vector of the execution unit running state; an association matching analysis processing module, configured to perform an association matching analysis based on historical data enhancement on the key feature aggregation vector of the execution unit operation state and the low-dimensional feature encoding vector of the control parameter to obtain a control parameter-operation state association matching encoding vector; The device control instruction generation module is used to generate environmental protection device control instructions based on the control parameter-operating state association matching coding vector.

[0015] The present invention also includes a device, which includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned environmental protection device control method when executing the computer program.

[0016] The present invention also includes a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned environmental protection equipment control method are implemented.

[0017] The beneficial effects of the environmental protection equipment control method, system, equipment and medium provided by the present invention are: The environmental protection equipment control method provided by the present invention provides clear constraints for the operation of environmental protection equipment by setting control parameters such as operating temperature thresholds, flow rate adjustment ranges, and pressure safety limits, ensuring that the equipment operates within a reasonable range and ensuring its safety and stability. The method collects and processes the operating status data of each execution unit of the controlled environmental protection equipment. Feature extraction and encoding of the control parameters are performed to generate low-dimensional feature encoding vectors. This operating status data is then subjected to layered filtering based on feature weights. This method effectively extracts key features, removes redundant information, improves the efficiency and accuracy of data processing, and enables subsequent analysis to be based on more valuable information.

[0018] This approach uses historical data-enhanced correlation matching analysis to analyze the aggregated key feature vectors of the execution unit's operating status and the low-dimensional feature encoding vectors of the control parameters. Leveraging historical data, this approach further explores the correlation between control parameters and operating status, making the correlation matching analysis more precise. This results in a control parameter-operating status correlation matching encoding vector that better reflects actual operating conditions. This encoding vector is then used to generate control instructions for environmental protection equipment. The decision-maker's control instruction generation module comprehensively considers multiple factors to generate more targeted and accurate control instructions, guiding precise operational adjustments for environmental protection equipment.

[0019] This control method adjusts control parameters in real time based on the equipment's actual operating status and historical experience, enabling environmental protection equipment to maintain efficient operation under varying operating conditions and external environments, improving the equipment's adaptability to complex operating conditions. Precise control instructions effectively enhance pollutant treatment while reducing energy consumption and improving equipment efficiency, achieving the dual goals of environmental protection and energy conservation. Furthermore, through effective data processing and analysis, this method enhances the system's intelligence, reduces manual intervention, lowers operating costs, and reduces maintenance difficulty, providing strong support for the automated and intelligent control of environmental protection equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments or descriptions of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0021] Figure 1 This is a working principle diagram of the environmental protection equipment control method of the present invention; Figure 2 Sub-flowchart for hierarchical filtering processing; Figure 3 Sub-flow graph optimized for fine-grained tuning; Figure 4 This is the module diagram of the environmental protection equipment control system. DETAILED DESCRIPTION

[0022] In the following description, specific details such as specific system structures and techniques are provided for purposes of illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present application with unnecessary detail.

[0023] In order to make the purpose, technical solutions and advantages of this application clearer, specific embodiments will be described below with reference to the accompanying drawings.

[0024] See also Figures 1-4 The present invention provides an environmental protection equipment control method, system, equipment and medium. The specific implementation steps are as follows: Control parameters are set, including an operating temperature threshold, a flow rate adjustment range, and a pressure safety limit.

[0025] The operating status data of each execution unit of the controlled environmental protection equipment is collected under the constraints of the control parameters to obtain a set of the operating status data of the execution units.

[0026] The control parameters are feature extracted and encoded to obtain a low-dimensional feature encoding vector of the control parameters. Specifically, parameters such as the operating temperature threshold, flow adjustment range, and pressure safety limit can be subjected to dimensionality reduction processing using a preset encoding model to remove redundant information, retain key feature dimensions, and form a low-dimensional feature encoding vector.

[0027] A hierarchical filtering process based on feature weights is performed on the set of execution unit running state data to obtain an execution unit running state key feature aggregation vector.

[0028] An association matching analysis based on historical data enhancement is performed on the key feature aggregation vector of the execution unit running state and the low-dimensional feature coding vector of the control parameter to obtain a control parameter-running state association matching coding vector.

[0029] Based on the control parameter-operating state association matching coding vector, an environmental protection equipment control instruction is generated, specifically: the control parameter-operating state association matching coding vector is input into a control instruction generation module based on a decision maker to obtain an equipment control instruction, and the equipment control instruction is used to instruct the operation adjustment operation of the environmental protection equipment.

[0030] Example 1: When performing hierarchical filtering processing based on feature weights on a set of execution unit operating status data to obtain an aggregate vector of key features of the execution unit operating status, the specific implementation method is as follows. Each execution unit operating status data in the set of execution unit operating status data passes through a filtering layer respectively. The filtering layer here can be a pre-set algorithm module with specific filtering rules, which functions to perform preliminary feature screening and encoding processing on each execution unit operating status data. In this process, each execution unit operating status data will be calculated through the filtering layer to remove some redundant or unimportant information while retaining representative features, thereby obtaining a set of execution unit operating status feature filtering coding vectors.

[0031] It is necessary to calculate the execution unit operation state pseudo-reference filter center representation vector based on the feature distribution density of the set of execution unit operation state feature filter coding vectors. This process requires that each execution unit operation state feature filter coding vector in the set of execution unit operation state feature filter coding vectors be input into the feature distribution density measurement module respectively. In the feature distribution density measurement module, each execution unit operation state feature filter coding vector is first normalized. The purpose of normalization is to convert the feature values ​​of different scales to the same scale for subsequent calculation and comparison. The specific normalization method can adopt a common normalization algorithm, such as dividing each element in the vector by the modulus of the vector to obtain a normalized execution unit operation state feature filter coding vector.

[0032] After obtaining the normalized execution unit operating state feature filter code vector, the square sum of each eigenvalue in the vector is calculated. This square sum is then divided by the feature dimension value of the normalized execution unit operating state feature filter code vector to obtain the execution unit operating state distribution density metric coefficient. In this way, each execution unit operating state feature filter code vector can be assigned a corresponding distribution density metric coefficient, and these coefficients constitute the set of execution unit operating state distribution density metric coefficients.

[0033] The set of execution unit operating state distribution density measurement coefficients is input into the feature density screening control unit. The feature density screening control unit, which is pre-configured with specific screening rules and weight calculation logic, processes the input distribution density measurement coefficients and assigns a corresponding execution unit operating state distribution density weight factor to each execution unit operating state feature filter code vector based on the size and distribution of each coefficient. This results in a set of execution unit operating state distribution density weight factors.

[0034] After obtaining a set of execution unit operating state distribution density weighting factors, a weighted sum of the set of execution unit operating state feature filter code vectors is calculated based on these weighting factors. Each execution unit operating state feature filter code vector is multiplied by its corresponding weighting factor, and all products are accumulated to obtain the execution unit operating state pseudo-baseline filtered center representation vector. This vector reflects, to a certain extent, the central tendency of the set of execution unit operating state feature filter code vectors.

[0035] The filtering adjustment direction of each execution unit operating state feature filter code vector in the set of execution unit operating state feature filter code vectors relative to the execution unit operating state pseudo-reference filter center representation vector is calculated, thereby obtaining a set of execution unit operating state filtering adjustment directions. The filtering adjustment direction can be determined by calculating the difference between each feature filter code vector and the pseudo-reference filter center representation vector. For example, the pseudo-reference filter center representation vector can be subtracted from the feature filter code vector, and the resulting difference vector is the filtering adjustment direction.

[0036] Based on the set of execution unit running state filter adjustment directions, the set of execution unit running state feature filter coding vectors is hierarchically filtered toward the execution unit running state pseudo-reference filter center representation vector. The hierarchical filtering process can be divided into multiple levels, and each layer adjusts and filters the feature filter coding vector according to the filter adjustment direction, gradually removing unimportant features that deviate from the center and retaining key features close to the center. After multiple layers of filtering processing, the execution unit running state key feature aggregation vector is finally aggregated. This vector contains the most critical feature information in the execution unit running state data, and provides an important basis for subsequent correlation matching analysis and control instruction generation. In the entire processing process, each step is strictly carried out in accordance with the preset algorithm and logic to ensure the accuracy and consistency of data processing.

[0037] Example 2: When performing historical data-enhanced association matching analysis on the key feature aggregation vector of the execution unit's operating state and the low-dimensional feature coding vector of the control parameters to obtain the control parameter-operating state association matching coding vector, the specific implementation method is as follows. Based on historical data, the key feature aggregation vector of the execution unit's operating state and the low-dimensional feature coding vector of the control parameters are optimized for feature association based on a matching mechanism to obtain a historical data optimized control parameter-operating state association matching feature matrix. Here, the key feature aggregation vector of the execution unit's operating state and the low-dimensional feature coding vector of the control parameters are first input into a fine-grained association analysis network. The network can adopt a deep learning architecture, such as a multi-layer perceptron or a convolutional neural network, and its internal nodes learn the association pattern between the two features through training data. The network receives two vectors as input, and after calculation and feature transformation by multiple layers of neurons, outputs a control parameter-operating state association matching feature matrix. Each element of the matrix represents the correlation strength of the feature dimension corresponding to the two vectors, and fine-grained association analysis between features is achieved through network operations.

[0038] The control parameter-operating state correlation matching feature matrix is ​​input into a historical data-based matching unit. The matching unit stores multiple sets of control parameter vectors, operating state vectors, and corresponding correlation matching matrices accumulated during historical operation. This historical data contains operational information about the equipment under different operating conditions and represents a summary of long-term operating experience. The matching unit compares the currently input correlation matching feature matrix with the historically stored matrices. Specifically, it calculates similarity metrics between the matrices, such as cosine similarity or Euclidean distance, to identify differences between the current and historical matrices. Based on the results of this difference analysis, the matching unit adjusts the current matrix, for example by modifying certain correlation strength values ​​and supplementing well-performing correlation patterns in the historical data. This results in an optimized control parameter-operating state correlation matching feature matrix for the historical data. This process effectively incorporates historical experience into the current correlation analysis, making the resulting matrix more consistent with the actual operating patterns of the equipment.

[0039] After obtaining the historical data optimized control parameter-operating state correlation matching feature matrix, it is necessary to perform feature adjustment optimization on the execution unit operating state key feature aggregation vector and the control parameter low-dimensional feature encoding vector respectively to obtain the optimized execution unit operating state key feature aggregation vector and the optimized control parameter low-dimensional feature encoding vector. For the execution unit operating state key feature aggregation vector, it is first linearly transformed. The linear transformation is achieved through a preset transformation matrix, the construction of which is based on the structure of the historical data optimized feature matrix, and aims to decompose the original vector into two parts: the first associated feature vector and the first adjusted feature vector. Among them, the first associated feature vector contains features that are highly correlated with the control parameters, and the first adjusted feature vector contains features that need to be further optimized.

[0040] Using the historical data optimized control parameter-operating status correlation matching feature matrix as the reference matrix, the first correlation feature vector, the first adjustment feature vector, and the reference matrix are input into the fine-grained adjustment module based on the memory network. The memory network stores successful cases of historical feature adjustments, each of which records the adjustment method and results under a specific feature combination. The fine-grained adjustment module uses the attention mechanism to retrieve the historical cases most relevant to the current feature adjustment needs from the memory network based on the strength of the feature correlation in the reference matrix. When processing the first correlation feature vector and the first adjustment feature vector, the module will redistribute the weights of different features and modify the feature values ​​based on the experience of historical cases. For example, it will increase the weights of certain key correlation features and reduce the influence of irrelevant features, thereby obtaining the optimized execution unit operating status key feature aggregation vector.

[0041] The low-dimensional feature encoding vector of the control parameters is subjected to feature adjustment optimization. First, a linear transformation is performed on it to obtain a second associated feature vector and a second adjusted feature vector. The process of linear transformation is similar to the transformation of the key feature aggregation vector of the running state of the execution unit. The original vector is also decomposed into a part with a high correlation with the running state characteristics and a part that needs to be adjusted through a preset matrix. Then, the control parameter-running state associated matching feature matrix is ​​optimized with the same historical data as the reference matrix, and the second associated feature vector, the second adjusted feature vector and the reference matrix are input into a fine-grained adjustment module based on a memory network. In the same way, the module performs fine-grained adjustment on the control parameter features based on historical cases to obtain an optimized low-dimensional feature encoding vector of the control parameters.

[0042] The bitwise product of the optimized execution unit operating state key feature aggregation vector and the optimized control parameter low-dimensional feature encoding vector is calculated to generate a control parameter-operating state correlation matching encoding vector. The bitwise product operation rule multiplies the elements of the two vectors at corresponding positions to obtain a new vector element. Through this operation, the feature correlation information of the two vectors is integrated into the new vector. Each element contains information about both the operating state characteristics and the control parameter characteristics, and the element value reflects the strength of the correlation between the two in that feature dimension. The control parameter-operating state correlation matching encoding vector generated in this way integrates the results of historical data-enhanced correlation analysis and the results of feature adjustment optimization. It can more accurately reflect the intrinsic connection between control parameters and operating state, providing a reliable basis for the subsequent generation of environmental protection equipment control instructions. Throughout the implementation process, the operational logic and data processing flow of each module are strictly based on preset algorithms and historical data, ensuring that each operation step has a clear logical support and data foundation, thereby ensuring the accuracy and effectiveness of the correlation matching analysis.

[0043] Example 3: When calculating the pseudo-reference filtered center representation vector of the execution unit running state, it involves inputting each vector in the set of the execution unit running state feature filter coding vector into the feature distribution density measurement module to obtain a set of execution unit running state distribution density measurement coefficients. The specific implementation method is as follows. First, it is necessary to normalize the execution unit running state feature filter coding vector to obtain a normalized execution unit running state feature filter coding vector. The normalization process is to eliminate the differences in scales between different feature vectors so that subsequent calculations can be performed under a unified standard. The specific operation is to calculate the sum of the squares of each eigenvalue of each execution unit running state feature filter coding vector, and then take the square root of the square sum to obtain the modulus of the vector. Then, each eigenvalue in the vector is divided by the modulus to obtain the normalized vector. After such processing, the modulus of each normalized execution unit running state feature filter coding vector is 1, which is convenient for subsequent distribution density calculations.

[0044] After obtaining the normalized execution unit operation state feature filter coding vector, it is necessary to calculate the execution unit operation state distribution density measurement coefficient of the vector. The specific calculation method is to first find the sum of the squares of each eigenvalue in the normalized vector. This sum of squares reflects the distribution concentration of the eigenvalues ​​in the vector space. Then, divide the sum of squares by the feature dimension value of the normalized execution unit operation state feature filter coding vector. The feature dimension value refers to the number of features contained in the vector. For example, a vector containing n features has a feature dimension value of n. By dividing the sum of squares by the feature dimension value, the distribution concentration degree can be converted into a relative measurement value, so that the distribution density between vectors of different dimensions is comparable. In this way, each normalized feature filter coding vector can obtain a corresponding execution unit operation state distribution density measurement coefficient, and all such coefficients constitute a set of execution unit operation state distribution density measurement coefficients.

[0045] Next, the set of execution unit operating state distribution density measurement coefficients is input into the feature density screening control unit. The feature density screening control unit is pre-configured with specific screening rules and weight assignment logic. This unit analyzes the input distribution density measurement coefficients and, based on the magnitude of each coefficient, determines the feature distribution density of the corresponding feature filter code vector. Generally speaking, vectors with larger distribution density measurement coefficients have more concentrated eigenvalues ​​in vector space and may contain more critical operating state information; vectors with smaller coefficients have more dispersed feature distributions and may contain more redundant information. Based on these coefficients, the feature density screening control unit assigns an execution unit operating state distribution density weight factor to each execution unit operating state feature filter code vector. The weight factor is positively correlated with the distribution density measurement coefficient: vectors with higher distribution density have larger weight factors, and vice versa. In this way, a set of execution unit operating state distribution density weight factors is obtained. Each weight factor corresponds to a feature filter code vector, indicating the importance of that vector in subsequent calculations.

[0046] After obtaining the set of execution unit running state distribution density weight factors, the weighted sum of the set of execution unit running state feature filter coding vectors is calculated based on these weight factors to obtain the execution unit running state pseudo-benchmark filter center representation vector. The specific calculation process is that for each execution unit running state feature filter coding vector, it is multiplied by the corresponding distribution density weight factor to obtain a weighted vector. Then, all weighted vectors are accumulated to obtain the final weighted sum vector, that is, the execution unit running state pseudo-benchmark filter center representation vector. This vector comprehensively considers the feature distribution density of all feature filter coding vectors. Vectors with high distribution density occupy a larger proportion in the weighted sum. Therefore, the pseudo-benchmark filter center representation vector can better reflect the central trend of the entire feature filter coding vector set.

[0047] In this process, the processing of each step needs to strictly follow the preset algorithms and logic. Normalization ensures the consistency of vector scales, the calculation of the distribution density measurement coefficient quantifies the concentration of feature distribution, the weight distribution of the feature density screening control unit reflects the difference in importance of different vectors, and the calculation of the weighted sum combines these factors to obtain a pseudo-benchmark filter center representation vector that can represent the center of the overall feature distribution. This calculation method can effectively extract a representative center vector from the set of feature filter coding vectors, providing a reliable benchmark for the subsequent calculation of the filtering adjustment direction of each feature filter coding vector relative to the center. During the entire implementation process, the data processing flow is clear, and the operations of each link have clear purposes and bases, ensuring the accuracy and effectiveness of the calculation results, laying the foundation for subsequent hierarchical filtering processing, and enabling more accurate extraction of key feature information from the execution unit operation status data.

[0048] Example 4: When performing association matching analysis on the aggregated key feature vectors of the execution unit's operating status and the low-dimensional feature encoding vectors of the control parameters, feature association optimization based on historical data is implemented through a fine-grained association analysis network and matching unit. The specific implementation method is as follows. For example, a certain industrial wastewater treatment and environmental protection equipment includes multiple execution units, such as a filtration unit, an aeration unit, and a sedimentation unit. The operating status data of each unit includes parameters such as flow rate, pressure, and temperature. The control parameters are set to include operating temperature thresholds, flow rate adjustment ranges, and pressure safety limits.

[0049] The aggregated vector of the key operational status features of the execution units and the low-dimensional feature encoding vectors of the control parameters are input into a fine-grained association analysis network. Assume that the aggregated vector of the key operational status features of the execution units includes the real-time flow characteristics of the filtration unit and the pressure characteristics of the aeration unit, and the low-dimensional feature encoding vectors of the control parameters include the set flow adjustment range characteristics and temperature threshold characteristics. The fine-grained association analysis network can adopt a multi-layer neural network structure. Its input layer receives these two vectors, and the hidden layer transforms and analyzes the features through weighted connections between neurons. For example, the first hidden layer of the network may perform weighted calculations on the flow characteristics and flow adjustment range characteristics, while the second hidden layer may perform nonlinear transformations combining the pressure characteristics and the temperature threshold characteristics. After multiple layers of calculation, the output is a control parameter-operational status correlation matching feature matrix. Each element in this matrix represents the degree of correlation between the corresponding control parameter feature and the operational status feature. For example, the strength of the correlation between the flow adjustment range feature and the filtration unit flow characteristic for a certain element in the matrix reflects the degree of correlation between the two.

[0050] The control parameter-operating status correlation matching feature matrix is ​​input into a matching unit based on historical data. The matching unit stores the wastewater treatment plant's operating data for the past year, including control parameter vectors, operating status vectors, and corresponding correlation matching matrices for different seasons and wastewater concentrations. For example, the historical data records the correlation matching between the set temperature threshold and the aeration unit temperature characteristics during high summer temperatures, when wastewater concentrations are high, and the corresponding treatment effects. The matching unit compares the currently generated correlation matching feature matrix with the historically stored matrices, specifically by calculating the similarity between the matrices to identify differences. For example, if the correlation strength between the temperature threshold characteristics and the aeration unit temperature characteristics in the current matrix is ​​0.6, while the historical data shows a typical correlation strength of 0.8 under similar operating conditions, the matching unit identifies this discrepancy and adjusts the value of this element in the current matrix to 0.75 based on the historical data. It also makes corresponding adjustments to other discrepant elements, thereby obtaining the historical data-optimized control parameter-operating status correlation matching feature matrix.

[0051] During this process, the fine-grained correlation analysis network is trained based on a large amount of data collected during normal equipment operation. Using a supervised learning approach, the historical correlation matching matrix serves as a label, and the network's weight parameters are adjusted to accurately learn the correlation patterns between control parameters and operating status characteristics. The matching unit's historical data, accumulated over the equipment's long-term operation, covers operational information under various operating conditions, ensuring the rationality and reliability of the current matrix adjustments.

[0052] Taking another scenario as an example, when the quality of the wastewater treated by the equipment changes, such as when oily wastewater is suddenly connected, the pressure characteristics of the filter unit in the key feature aggregation vector of the execution unit's operating status will change significantly, and the flow regulation range characteristics in the low-dimensional feature encoding vector of the control parameter may need to be adjusted accordingly. At this time, the fine-grained association analysis network will generate a new association matching feature matrix based on the new input vector, in which the correlation strength between the filter unit pressure characteristics and the flow regulation range characteristics may be different from the normal operating conditions in the historical data. When comparing historical data, the matching unit will retrieve the historical association matching matrix when the water quality changes, find the difference in the correlation strength in the current matrix, and optimize it based on historical experience, so that the adjusted matrix is ​​more in line with actual processing needs.

[0053] The synergistic effect of the fine-grained correlation analysis network and the matching unit enables feature correlation optimization based on historical data. The fine-grained correlation analysis network extracts the correlation relationships between features from the current input vector to generate an initial correlation matching matrix. The matching unit, drawing on the experience accumulated from historical data, corrects and optimizes this initial matrix, compensating for any deficiencies in the current data. The resulting historical data-optimized control parameter-operating state correlation matching feature matrix more accurately reflects the actual correlation between control parameters and operating states, providing a more reliable basis for subsequent feature adjustment optimization and control instruction generation.

[0054] Example 5: When optimizing the control parameter-operating state correlation matching feature matrix based on historical data for feature adjustment optimization, the implementation method is described using a chemical waste gas treatment and environmental protection equipment as an example. The equipment's execution units include a catalytic combustion unit and an adsorption unit. The control parameters are set as a catalytic temperature threshold of 280-320°C, an adsorption flow rate adjustment range of 1500-2000 m³ / h, and a pressure safety limit of 80 kPa. Assume that during a certain operating period, the key feature aggregation vector of the execution unit's operating state includes the real-time temperature feature of the catalytic combustion unit (295°C) and the real-time flow rate feature of the adsorption unit (1800 m³ / h). The low-dimensional feature encoding vector of the control parameter includes the catalytic temperature threshold feature (with the interval of 280-320°C mapped as the feature vector) and the flow rate adjustment range feature (with the interval of 1500-2000 m³ / h mapped as the feature vector).

[0055] A linear transformation is performed on the aggregate vector of the key features of the execution unit's operating status. This linear transformation is achieved through a preset transformation matrix, the construction of which is based on the structure of the historical data-optimized control parameter-operating status correlation matching feature matrix. For example, the historical data optimization feature matrix shows that the catalytic temperature feature has a high correlation with the catalytic temperature threshold feature, while the adsorption flow feature has a second-lowest correlation with the flow regulation range feature. Based on this, the linear transformation decomposes the aggregate vector of the key features of the execution unit's operating status into a first correlation feature vector and a first adjustment feature vector. The first correlation feature vector primarily contains features with a high correlation with the control parameters, such as the dimension corresponding to the real-time feature of the catalytic temperature (295°C). The first adjustment feature vector contains features that require further optimization, such as the dimension corresponding to the real-time feature of the adsorption flow (1800 m³ / h).

[0056] Using the historical data-optimized control parameter-operating state correlation matching feature matrix as a reference matrix, the first correlation feature vector, the first adjustment feature vector, and the reference matrix are input into a fine-grained adjustment module based on a memory network. The memory network stores the exhaust gas treatment equipment's operational records from the past two years. For example, when the real-time catalytic temperature feature was 290°C, historical adjustment experience indicated that the weight of the temperature threshold feature should be increased by 0.15 to ensure catalytic efficiency. When the real-time adsorption flow rate feature was 1900 m³ / h, historical examples revealed that flow rate fluctuations had caused adsorption efficiency to decrease, so the corresponding adjustment method was to modify certain dimensions of the flow regulation range feature. The fine-grained adjustment module uses an attention mechanism to retrieve the historical cases most relevant to the current catalytic temperature feature from the memory network based on the correlation strength between the catalytic temperature feature and the temperature threshold feature in the reference matrix (assuming the value of the element at that position in the reference matrix is ​​0.78). For example, adjustment experience for temperatures between 290°C and 300°C was used. Based on this historical case, the module adjusts the weight of the catalytic temperature feature dimension in the first associated feature vector from the initial 0.6 to 0.7. At the same time, it corrects the adsorption flow feature dimension in the first adjusted feature vector. Referring to the adjustment method when the flow rate was 1900 m³ / h in history, the module fine-tunes the feature value of this dimension, for example, adjusting its mapping value from 0.8 to 0.85, thereby obtaining the optimized key feature aggregation vector of the execution unit operation status.

[0057] The low-dimensional feature encoding vector of the control parameters is subjected to feature adjustment and optimization. A linear transformation is first performed to obtain a second correlation feature vector and a second adjustment feature vector. This linear transformation, based on the same historical data optimization feature matrix, decomposes the low-dimensional feature encoding vector of the control parameters into a portion highly correlated with the operating status characteristics and a portion requiring adjustment. For example, the second correlation feature vector includes dimensions of the catalytic temperature threshold feature that are highly correlated with the catalytic combustion unit temperature characteristics, while the second adjustment feature vector includes dimensions of the flow regulation range feature that require optimization in their correlation with the adsorption unit flow characteristics.

[0058] The control parameter-operation state association matching feature matrix optimized with the same historical data is taken as a reference matrix, and the second association feature vector, the second adjustment feature vector, and the reference matrix are input into the fine-grained adjustment module based on the memory network. According to the association strength (assuming the element value is 0.65) of the flow adjustment range feature and the adsorption flow feature in the reference matrix, the module retrieves the historical adjustment cases of the flow adjustment range at 1800-2000 m³ / h in the memory network. For example, when the association strength of the flow adjustment range feature and the real-time flow feature is 0.6, the historical adjustment increases the weight of a certain sub-dimension of the flow adjustment range feature by 0.1. The module adjusts the weight of the flow adjustment range feature related dimension in the second association feature vector according to the case, and corrects other dimensions in the second adjustment feature vector to obtain the optimized low-dimensional feature encoding vector of the control parameter.

[0059] The bitwise product between the optimized execution unit operation state key feature aggregation vector and the optimized low-dimensional feature encoding vector of the control parameter is calculated to obtain the control parameter-operation state association matching encoding vector. For example, the value of the catalytic temperature feature dimension in the optimized operation state vector is 0.7, and the value of the temperature threshold feature corresponding dimension in the optimized control parameter vector is 0.8. After bitwise multiplication, the value of this dimension is 0.56; the value of the adsorption flow feature dimension in the operation state vector is 0.85, and the value of the flow adjustment range feature corresponding dimension in the control parameter vector is 0.7. After bitwise multiplication, the value of this dimension is 0.595. Through this operation, the value of each dimension of the generated new vector integrates the association information of the operation state feature and the control parameter feature, reflecting the matching degree of the two in this dimension.

[0060] In another scenario, if the device encounters a sudden working condition, such as the temperature of the catalytic combustion unit suddenly rising to 310°C due to changes in waste gas composition, the value of the temperature feature dimension in the execution unit operation state key feature aggregation vector changes significantly. At this time, the linear conversion will re-decompose the vector, and the fine-grained adjustment module retrieves historical adjustment cases under high-temperature working conditions from the memory network, such as reducing the weight of a certain dimension of the temperature threshold feature by 0.1 when the temperature is 310°C to trigger the warning mechanism. The module adjusts the features according to the case, so that the optimized vector is more suitable for the current working condition, and the association matching encoding vector generated by bitwise multiplication can accurately reflect the association between the control parameter and the operation state under high-temperature conditions, providing a basis for generating control instructions to adjust the catalytic temperature subsequently.

[0061] Throughout the implementation, the linear transformation matrix parameters are derived from historical data collected from the device's long-term operation. The memory network examples are derived from device maintenance records and operation logs, ensuring that each adjustment step is supported by real-world data. The fine-grained adjustment module's attention mechanism dynamically matches the most relevant historical experience, avoiding the limitations of subjective parameter settings and ensuring that feature adjustments are more consistent with the device's actual operating patterns.

[0062] The above are only specific embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and such modifications or substitutions should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A method for controlling environmental protection equipment, characterized in that: include: Setting control parameters, including operating temperature threshold, flow rate adjustment range, and pressure safety limit; Collecting the operating status data of each execution unit of the controlled environmental protection equipment under the constraints of the control parameters to obtain a set of the operating status data of the execution units; Performing feature extraction and encoding on the control parameter to obtain a low-dimensional feature encoding vector of the control parameter; Performing a hierarchical filtering process based on feature weights on the set of execution unit running state data to obtain a key feature aggregation vector of the execution unit running state; Performing an association matching analysis based on historical data enhancement on the key feature aggregation vector of the execution unit operation state and the low-dimensional feature coding vector of the control parameter to obtain a control parameter-operation state association matching coding vector; Based on the control parameter-operating state association matching coding vector, an environmental protection equipment control instruction is generated, specifically: the control parameter-operating state association matching coding vector is input into a control instruction generation module based on a decision maker to obtain an equipment control instruction, and the equipment control instruction is used to instruct the operation adjustment operation of the environmental protection equipment.

2. The environmental protection equipment control method according to claim 1, characterized in that: Performing a hierarchical filtering process based on feature weights on the set of execution unit running state data to obtain a key feature aggregation vector of the execution unit running state, including: Passing each execution unit operation status data in the set of execution unit operation status data through a filtering layer to obtain a set of execution unit operation status feature filtering coding vectors; Calculating a pseudo-reference filtered center representation vector of the execution unit running state based on a feature distribution density of a set of the execution unit running state feature filtering code vectors; Calculating a filtering adjustment direction of each execution unit operation state feature filtering code vector in the set of the execution unit operation state feature filtering code vectors relative to the execution unit operation state pseudo-reference filtering center representation vector to obtain a set of execution unit operation state filtering adjustment directions; Based on the set of execution unit operation state filtering adjustment directions, the set of execution unit operation state feature filtering encoding vectors is hierarchically filtered toward the execution unit operation state pseudo-reference filtering center representation vector to obtain the execution unit operation state key feature aggregation vector.

3. The environmental protection equipment control method according to claim 2, characterized in that: Calculating the execution unit running state pseudo-reference filtered center representation vector based on the feature distribution density of the set of the execution unit running state feature filtering code vectors includes: Inputting each execution unit operation state feature filter code vector in the set of execution unit operation state feature filter code vectors into a feature distribution density measurement module to obtain a set of execution unit operation state distribution density measurement coefficients; Inputting the set of execution unit operation state distribution density measurement coefficients into a feature density screening control unit to obtain a set of execution unit operation state distribution density weight factors; Based on the set of execution unit operation state distribution density weight factors, a weighted sum of the set of execution unit operation state feature filter coding vectors is calculated to obtain the execution unit operation state pseudo-reference filter center representation vector.

4. The environmental protection equipment control method according to claim 3, characterized in that: Inputting each execution unit operation state feature filter code vector in the set of the execution unit operation state feature filter code vectors into a feature distribution density measurement module to obtain a set of execution unit operation state distribution density measurement coefficients, including: Normalizing the execution unit operation state feature filtering code vector to obtain a normalized execution unit operation state feature filtering code vector; The execution unit operation state distribution density measurement coefficient is obtained by calculating the sum of squares of each eigenvalue in the normalized execution unit operation state feature filtering code vector and dividing it by the feature dimension value of the normalized execution unit operation state feature filtering code vector.

5. The environmental protection equipment control method according to claim 4, characterized in that: Performing an association matching analysis based on historical data enhancement on the key feature aggregation vector of the execution unit operation state and the low-dimensional feature encoding vector of the control parameter to obtain a control parameter-operation state association matching encoding vector, including: Based on historical data, performing feature association optimization based on a matching mechanism on the key feature aggregation vector of the execution unit operation state and the low-dimensional feature encoding vector of the control parameter to obtain a historical data optimized control parameter-operation state association matching feature matrix; Optimizing the control parameter-operation state association matching feature matrix based on the historical data, performing feature adjustment optimization on the execution unit operation state key feature aggregation vector and the control parameter low-dimensional feature encoding vector to obtain an optimized execution unit operation state key feature aggregation vector and an optimized control parameter low-dimensional feature encoding vector; The bitwise product between the optimized execution unit running state key feature aggregation vector and the optimized control parameter low-dimensional feature encoding vector is calculated to obtain the control parameter-running state association matching encoding vector.

6. The environmental protection equipment control method according to claim 5, characterized in that: Based on historical data, feature association optimization is performed on the key feature aggregation vector of the execution unit operation state and the low-dimensional feature encoding vector of the control parameter based on a matching mechanism to obtain a historical data optimized control parameter-operation state association matching feature matrix, including: Inputting the key feature aggregation vector of the execution unit operation state and the low-dimensional feature encoding vector of the control parameter into a fine-grained association analysis network to obtain a control parameter-operation state association matching feature matrix; The control parameter-operating state association matching characteristic matrix is ​​input into a matching unit based on historical data to obtain the historical data optimized control parameter-operating state association matching characteristic matrix.

7. The environmental protection equipment control method according to claim 6, characterized in that: Optimizing the control parameter-operating state association matching feature matrix based on the historical data, and performing feature adjustment optimization on the execution unit operating state key feature aggregation vector and the control parameter low-dimensional feature encoding vector to obtain an optimized execution unit operating state key feature aggregation vector and an optimized control parameter low-dimensional feature encoding vector, including: performing a linear transformation on the execution unit operating state key feature aggregation vector to obtain a first correlation feature vector and a first adjustment feature vector, using the historical data optimized control parameter-operating state correlation matching feature matrix as a reference matrix, inputting the first correlation feature vector, the first adjustment feature vector, and the reference matrix into a fine-grained adjustment module based on a memory network to obtain the optimized execution unit operating state key feature aggregation vector; The control parameter low-dimensional feature encoding vector is linearly transformed to obtain a second associated feature vector and a second adjusted feature vector, and the historical data optimized control parameter-operating state associated matching feature matrix is ​​used as a reference matrix. The second associated feature vector, the second adjusted feature vector and the reference matrix are input into the memory network-based fine-grained adjustment module to obtain the optimized control parameter low-dimensional feature encoding vector.

8. An environmental protection equipment control system, used to implement the environmental protection equipment control method according to any one of claims 1 to 7, characterized in that: include: A control parameter setting module is used to set control parameters, including an operating temperature threshold, a flow rate adjustment range, and a pressure safety limit; An operation status data collection module, configured to collect the operation status data of each execution unit of the controlled environmental protection equipment under the constraints of the control parameters to obtain a set of execution unit operation status data; A control parameter encoding module, configured to perform feature extraction and encoding on the control parameter to obtain a low-dimensional feature encoding vector of the control parameter; a feature hierarchical filtering processing module, configured to perform a hierarchical filtering process on the set of the execution unit running state data based on feature weights to obtain a key feature aggregation vector of the execution unit running state; an association matching analysis processing module, configured to perform an association matching analysis based on historical data enhancement on the key feature aggregation vector of the execution unit operation state and the low-dimensional feature encoding vector of the control parameter to obtain a control parameter-operation state association matching encoding vector; The device control instruction generation module is used to generate environmental protection device control instructions based on the control parameter-operating state association matching coding vector.

9. A device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the environmental protection equipment control method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the environmental protection equipment control method according to any one of claims 1 to 7 are implemented.

Images