Sandstone crushing production process adaptive control method and system

Abstract

The application discloses a sandstone crushing production process self-adaptive control method and system, and belongs to the technical field of intelligent control of industrial production. The method comprises production data acquisition, blockage risk assessment, production control optimization and self-adaptive control execution. The application adopts a sandstone crushing blockage risk assessment method based on hierarchical fusion of space-time characteristics, performs self-adaptive identification on equipment load abnormalities under the conditions of sudden change of feeding and working condition switching, realizes early assessment of short-term evolution trend of the blockage risk, and provides a forward-looking risk constraint basis for subsequent production control optimization. A multi-agent collaborative game production control optimization method based on risk perception is adopted to coordinate and adjust the control behaviors of multiple devices such as feeding, crushing and screening under the condition of blockage risk constraint, so that the control strategies of each device dynamically converge to a safe, stable and efficient operating state with the change of the risk, thereby ensuring the continuity of production, reducing the probability of blockage and improving the overall operating effect.

Description

Technical Field

[0001] This invention belongs to the field of intelligent control technology for industrial production, specifically referring to an adaptive control method and system for sand and gravel crushing production process. Background Technology

[0002] Adaptive control of the sand and gravel crushing production process, combined with risk perception and multi-agent collaborative control technology, dynamically assesses the risks and performance indicators in the sand and gravel crushing production process through real-time perception and analysis of multi-source operating status information. Based on this, it adaptively adjusts the operating parameters of key equipment and collaborative control strategies, which helps to improve the stability and robustness of the production process, reduce energy consumption and the probability of equipment failure, and ensure the safe, efficient and economical operation of the production line under complex working conditions.

[0003] However, in the existing adaptive control process of sand and gravel crushing production, there are several problems. The operating load of equipment is significantly affected by the fluctuation of feed and the changes in material properties. Blockage often develops gradually, and existing methods rely on the current or vibration threshold at a single moment for judgment. It is difficult to distinguish between short-term operating condition fluctuations and the actual blockage evolution process in a timely manner. This leads to delayed identification of blockage risk and frequent false alarms. Consequently, control adjustments often occur after blockage has already formed or is close to forming. Furthermore, each key piece of equipment is usually adjusted independently according to its own operating conditions. The overall control method is mainly based on single-machine optimization, which makes it difficult to form a collaborative optimization mechanism for the entire production line. When the risk of blockage increases, the production system cannot effectively balance output, energy consumption and operational safety. This can easily lead to technical problems such as uncoordinated upstream feeding and downstream processing, frequent speed limits or shutdowns of local equipment, and large fluctuations in overall production efficiency. Summary of the Invention

[0004] To address the above issues and overcome the shortcomings of existing technologies, this invention provides an adaptive control method and system for sand and gravel crushing production processes. It creatively employs a risk assessment method for material blockage in sand and gravel crushing based on hierarchical fusion of spatiotemporal features. This method enables adaptive identification of abnormal equipment loads under conditions of sudden feed changes and operating mode switching. Furthermore, it provides advance assessment of the short-term evolution trend of material blockage risk, thus offering a forward-looking risk constraint basis for subsequent production control optimization. The invention also creatively employs a multi-agent collaborative game-based production control optimization method based on risk perception. This method achieves coordinated adjustment of the control behaviors of multiple equipment, including feeding, crushing, and screening, under material blockage risk constraints. This allows the control strategies of each equipment to dynamically converge to a safe, stable, and efficient operating state as the risk changes, thereby reducing the probability of material blockage and improving overall operational efficiency while ensuring production continuity.

[0005] The technical solution adopted by this invention is as follows: This invention provides an adaptive control method for a sand and gravel crushing production process, which includes the following steps:

[0006] Step S1: Production data collection;

[0007] Step S2: Material blockage risk assessment;

[0008] Step S3: Production control optimization;

[0009] Step S4: Adaptive control execution.

[0010] Further, in step S1, the production data acquisition specifically involves acquiring raw signal data through current sensors, vibration sensors, temperature sensors, feed rate detection devices, and discharge port pressure detection devices installed on the feeder, crusher, and screening machine, and performing time synchronization, multi-source data cleaning, and feature standardization processing to obtain a multi-dimensional production dataset.

[0011] Further, in step S2, the blockage risk assessment specifically involves using a sand and gravel crushing blockage risk assessment method based on spatiotemporal feature hierarchical fusion, according to a multidimensional production dataset, to obtain comprehensive blockage risk assessment information, including the following steps:

[0012] Step S21: Load deviation construction, specifically, by performing dynamic baseline modeling on the main motor current signal and equipment vibration signal, the dynamic mean and dynamic dispersion of the corresponding signals are obtained, and the load signal at the current time step is standardized based on the dynamic mean and dynamic dispersion to construct a multi-source load deviation vector and calculate the instantaneous load anomaly intensity.

[0013] The dynamic baseline modeling specifically involves statistically modeling the main motor current signal and the equipment vibration signal using an exponential sliding update method, respectively, to obtain the dynamic mean and dynamic dispersion that are adaptively adjusted with changes in operating conditions in real time, so as to characterize the normal load baseline level under the current operating state.

[0014] Step S22: State trend prediction, used to assess the evolution trend of material blockage risk in the near future. Specifically, it involves constructing a multivariate historical time series as input data, using a long short-term memory network to predict the main motor load and equipment vibration status, and calculating the predicted trend risk intensity based on the degree of deviation between the prediction results and the current dynamic baseline.

[0015] The multivariate historical time series includes feed rate, main motor current, equipment vibration, equipment temperature, and discharge port pressure;

[0016] The prediction results include a main motor current prediction sequence and a device vibration prediction sequence;

[0017] Step S23: Hierarchical risk fusion, specifically, the instantaneous load anomaly intensity is mapped to instantaneous load anomaly risk through nonlinear mapping, a predicted trend risk gating factor is introduced to dynamically adjust the predicted risk weight, and combined with historical risk accumulation terms, multi-level risk information is weighted and fused to obtain the comprehensive risk of material blockage.

[0018] The multi-level risk information includes instantaneous load anomaly risk, predicted trend risk intensity, and historical risk accumulation.

[0019] Step S24: Risk level determination, used to classify the risk of material blockage, specifically by mapping the comprehensive risk of material blockage to a preset risk range through nonlinear mapping to obtain the risk level of material blockage;

[0020] Step S25: Generate comprehensive information. Specifically, by executing steps S21 to S24, the blockage risk of the feeder, crusher and screening machine is evaluated respectively. The instantaneous load abnormality risk, predicted trend risk intensity, comprehensive blockage risk and blockage risk level of each equipment are integrated to obtain comprehensive information on blockage risk assessment.

[0021] Further, in step S3, the production control optimization is used to generate a collaborative control strategy that balances production efficiency, product quality, energy consumption, and operational safety under the constraint of material blockage risk. Specifically, based on the multi-dimensional production dataset and comprehensive information on material blockage risk assessment, an adaptive production control strategy package is obtained using a risk-aware multi-agent collaborative game production control optimization method, including the following steps:

[0022] Step S31: Multi-agent construction, used to build a collaborative decision-making environment for the sand and gravel crushing production process under the condition of material blockage risk perception. Specifically, by constructing the feeder, crusher and screening machine in the sand and gravel crushing production line as independent agents, and constructing agent state vectors for each agent, the multi-agent collaborative control environment is modeled to obtain the multi-agent collaborative decision-making environment model.

[0023] The agent state vector includes the equipment operating status, the comprehensive material blockage risk corresponding to each equipment, and the system-level comprehensive material blockage risk.

[0024] Step S32: Control action space constraint, which is used to actively suppress the control behavior of each device when the risk of material blockage increases. Specifically, it is to set corresponding control action variables for each agent and perform risk modulation on the control action space of each agent based on the comprehensive information of material blockage risk assessment, and adaptively shrink the control action range to obtain the risk-constrained control action space.

[0025] Step S33: Risk utility construction, used to achieve adaptive risk trade-off between overall system operating efficiency and equipment operating safety. Specifically, it involves constructing a global objective function and dynamically adjusting the objective weight based on the system-level comprehensive material blockage risk, while constructing local utility functions for each agent to obtain a game utility model.

[0026] The global objective function specifically uses the output, energy consumption, product particle size deviation, and system-level comprehensive material blockage risk of the sand and gravel crushing production system as indicators.

[0027] Step S34: Hierarchical game solution, specifically, based on the preset initial control strategy, the operation results of the sand and gravel crushing production system under different control strategy combinations are predicted by the process-level digital twin prediction model to obtain the system operation prediction output, and based on the system operation prediction output, the control actions of each agent are iteratively optimized by the hierarchical game update mechanism to obtain the optimal collaborative control strategy set of multiple agents.

[0028] The system operation prediction output includes system output prediction, system energy consumption prediction, product particle size deviation prediction, and system-level comprehensive material blockage risk prediction.

[0029] The hierarchical game update mechanism is specifically as follows: while keeping the control actions of other agents unchanged, each agent updates its control actions one by one according to the corresponding local utility function in a preset order, so as to gradually approach the stable state of the overall system utility.

[0030] Step S35: Strategy package generation, specifically, by integrating the optimal collaborative control strategy set of multi-agents, an adaptive production control strategy package containing the optimal control instructions of each device is generated, and the corresponding applicable material blockage risk range, expected operating effect evaluation information, and risk over-limit retreat conditions are associated in the adaptive production control strategy package.

[0031] Furthermore, in step S4, the adaptive control execution specifically involves adaptively controlling the sand and gravel crushing production process based on the adaptive production control strategy package, continuously monitoring key operating indicators and changes in material blockage risk during the control execution process, and achieving adaptive stable operation and material blockage risk suppression of the sand and gravel crushing production process by adopting strategy effect feedback and risk-triggered rollback control operations, thereby generating an adaptive control execution report.

[0032] The present invention provides an adaptive control system for a sand and gravel crushing production process, comprising a data acquisition module, a risk assessment module, a control optimization module, and a control execution module;

[0033] The data acquisition module is used for production data acquisition. Through production data acquisition, a multidimensional production dataset is obtained, and the multidimensional production dataset is sent to the risk assessment module and the control optimization module.

[0034] The risk assessment module is used for material blockage risk assessment. Through the material blockage risk assessment, comprehensive material blockage risk assessment information is obtained, and the comprehensive material blockage risk assessment information is sent to the control optimization module.

[0035] The control optimization module is used for production control optimization. Through production control optimization, an adaptive production control strategy package is obtained, and the adaptive production control strategy package is sent to the control execution module.

[0036] The control execution module is used for adaptive control execution and to obtain an adaptive control execution report.

[0037] The beneficial effects achieved by the present invention using the above solution are as follows:

[0038] (1) In the existing adaptive control process of sand and gravel crushing production, the equipment operating load is significantly affected by the fluctuation of feed and the change of material properties. The blockage often shows a gradual development characteristic. However, the existing methods mostly rely on the current or vibration threshold at a single moment to make judgments, which makes it difficult to distinguish between short-term operating condition fluctuations and the actual blockage evolution process in time. This leads to the delayed identification of blockage risk and frequent false alarms, which in turn causes the control adjustment to often occur after the blockage has formed or is close to forming. This solution creatively adopts a sand and gravel crushing blockage risk assessment method based on the hierarchical fusion of spatiotemporal features. It realizes the adaptive identification of equipment load anomalies under the conditions of sudden feed changes and operating condition switching. On this basis, it conducts an early assessment of the short-term evolution trend of blockage risk, thereby providing a forward-looking risk constraint basis for subsequent production control optimization.

[0039] (2) In the existing adaptive control process of sand and gravel crushing production, each key equipment is usually adjusted independently according to its own working conditions. The overall control mode is mainly based on single-machine optimization, which makes it difficult to form a collaborative optimization mechanism for the entire production line. When the risk of material blockage increases, the production system is difficult to achieve an effective balance between output, energy consumption and operational safety, which can easily lead to the technical problems of uncoordinated upstream feeding and downstream processing, frequent speed limits or shutdowns of local equipment, and large fluctuations in overall production efficiency. This solution creatively adopts a multi-agent collaborative game production control optimization method based on risk perception. It realizes the coordinated adjustment of the control behavior of multiple equipment such as feeding, crushing and screening under the constraint of material blockage risk. This enables the control strategies of each equipment to dynamically converge to a safe, stable and efficient operating state as the risk changes, thereby reducing the probability of material blockage and improving the overall operating effect while ensuring production continuity. Attached Figure Description

[0040] Figure 1 A flowchart illustrating an adaptive control method for a sand and gravel crushing production process provided by the present invention;

[0041] Figure 2 A schematic diagram of a module of an adaptive control system for a sand and gravel crushing production process provided by the present invention;

[0042] Figure 3 This is a flowchart illustrating step S2;

[0043] Figure 4 This is a flowchart illustrating step S3.

[0044] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof. Detailed Implementation

[0045] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0046] In the description of this invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0047] Example 1, see Figure 1 The present invention provides an adaptive control method for a sand and gravel crushing production process, the method comprising the following steps:

[0048] Step S1: Production data collection;

[0049] Step S2: Material blockage risk assessment;

[0050] Step S3: Production control optimization;

[0051] Step S4: Adaptive control execution.

[0052] Example 2, see Figure 1This embodiment is based on the above embodiment. In step S1, the production data acquisition is used to obtain multi-source operating status information of each key equipment in the sand and gravel crushing production process. Specifically, it involves acquiring raw signal data through current sensors, vibration sensors, temperature sensors, feed rate detection devices, and discharge port pressure detection devices installed on the feeder, crusher, and screening machine, and performing time synchronization, multi-source data cleaning, and feature standardization processing to obtain a multi-dimensional production dataset characterizing the real-time operating status of each piece of equipment.

[0053] The time synchronization is specifically achieved by uniformly adding timestamp identifiers to the data collected by various sensors, and resampling or interpolating and aligning data with different sampling frequencies at a preset sampling period, so that the operating data from different devices and different sensors remain consistent on the same time axis.

[0054] The multi-source data cleaning specifically involves performing outlier detection, noise filtering, and missing data processing on the collected raw signal data to remove invalid data caused by sensor jitter, communication anomalies, or transient interference, and to compensate for missing data using forward filling, linear interpolation, or moving average methods.

[0055] The feature standardization specifically involves scaling the cleaned signal data to map operating parameters of different dimensions to a unified numerical range, thereby eliminating the impact of dimensional differences on subsequent blockage risk assessment and control strategy generation.

[0056] Example 3, see Figure 1 and Figure 3 This embodiment is based on the above embodiment. In step S2, the blockage risk assessment is used to perceive the blockage risk of each piece of equipment in real time and predict its short-term trend during the sand and gravel crushing production process. Specifically, based on a multi-dimensional production dataset, a sand and gravel crushing blockage risk assessment method based on spatiotemporal feature hierarchical fusion is used to obtain comprehensive blockage risk assessment information, including the following steps:

[0057] Step S21: Load deviation construction, used to form an adaptive load deviation characterization under conditions of sudden feeding changes or working condition switching. Specifically, by performing dynamic baseline modeling on the main motor current signal and equipment vibration signal, the dynamic mean and dynamic dispersion of the corresponding signals are obtained. Based on the dynamic mean and dynamic dispersion, the load signal of the current time step is standardized to construct a multi-source load deviation vector and calculate the instantaneous load anomaly intensity.

[0058] The dynamic baseline modeling specifically involves statistically modeling the main motor current signal and the equipment vibration signal using an exponential sliding update method, respectively, to obtain the dynamic mean and dynamic dispersion that are adaptively adjusted with changes in operating conditions in real time, so as to characterize the normal load baseline level under the current operating state.

[0059] The formula for calculating the multi-source load deviation vector is:

[0060] ;

[0061] In the formula, This is the multi-source load deviation vector at time step t, where t is the first index of the time step, used to represent the current time step. t It is the main motor current signal at time step t. It is the dynamic average value of the main motor current. It is the dynamic dispersion of the main motor current, V t It is the equipment vibration signal at time step t. It is the dynamic average value of the equipment vibration signal. It is the dynamic dispersion of the equipment vibration signal;

[0062] The formula for calculating the instantaneous load anomaly intensity is as follows:

[0063] ;

[0064] In the formula, S t It represents the instantaneous load anomaly intensity at time step t, and ||·||2 is the L2 norm operator;

[0065] Step S22: State trend prediction, used to assess the evolution trend of material blockage risk in the near future. Specifically, it involves constructing a multivariate historical time series as input data, using a long short-term memory network to predict the main motor load and equipment vibration status, and calculating the predicted trend risk intensity based on the degree of deviation between the prediction results and the current dynamic baseline.

[0066] The multivariate historical time series includes feed rate, main motor current, equipment vibration, equipment temperature, and discharge port pressure;

[0067] The prediction results include a main motor current prediction sequence and a device vibration prediction sequence;

[0068] The formula for calculating the intensity of the predicted trend risk is as follows:

[0069] ;

[0070] In the formula, R trend It predicts the intensity of trend risk, max(·) is the function to find the maximum value, D I It is the deviation of current prediction, D V It is the deviation of vibration prediction;

[0071] The formulas for calculating the current prediction deviation and vibration prediction deviation are as follows:

[0072] ;

[0073] ;

[0074] In the formula, k is the prediction time step index, and H is the prediction step size. This is the predicted value of the main motor current at the k-th prediction time step. It is the predicted value of equipment vibration at the k-th prediction time step;

[0075] Step S23: Hierarchical risk fusion, used to uniformly quantify the instantaneous load anomaly risk, the intensity of predicted trend risk, and the historical material blockage accumulation risk. Specifically, the instantaneous load anomaly intensity is mapped to instantaneous load anomaly risk through nonlinear mapping, a predicted trend risk gating factor is introduced to dynamically adjust the predicted risk weight, and combined with the historical risk accumulation term, the multi-level risk information is weighted and fused to obtain the comprehensive material blockage risk.

[0076] The formula for calculating the predicted trend risk gating factor is as follows:

[0077] ;

[0078] In the formula, G t It is a trend risk gating factor. It is the sigmoid function, and k is the gating sensitivity coefficient. It is the threshold for triggering trend risk;

[0079] The historical risk accumulation item is specifically based on the intensity of instantaneous load anomalies, and is obtained by recursively accumulating the instantaneous load anomaly states over time. The calculation formula is as follows:

[0080] ;

[0081] In the formula, H t It represents the historical risk accumulation from the initial time step to the current time step, where i is the second index of the time step. It is the time decay weight, I(·) is the indicator function, and S i S is the instantaneous load anomaly intensity at time step i. th It is the threshold for instantaneous anomaly detection;

[0082] The multi-level risk information includes instantaneous load anomaly risk, predicted trend risk intensity, and historical risk accumulation.

[0083] The formula for calculating the overall risk of material blockage is as follows:

[0084] ;

[0085] In the formula, R equ It is a comprehensive risk of material blockage. It is the first weight in the weighted fusion, R inst It is a risk of sudden load anomalies. It is the second weight of the weighted fusion. It is a weighted fusion third weight, which satisfies ;

[0086] Step S24: Risk level determination, used to classify the risk of material blockage, specifically by mapping the comprehensive risk of material blockage to a preset risk range through nonlinear mapping to obtain the risk level of material blockage;

[0087] The preset risk range is specifically set by statistically modeling the comprehensive risk of material blockage under historical normal working conditions, material blockage evolution conditions, and material blockage occurrence conditions, and adaptively using quantile division, threshold optimization, or unsupervised clustering methods according to the distribution characteristics of the comprehensive risk of material blockage.

[0088] Step S25: Generate comprehensive information. Specifically, by executing steps S21 to S24, the blockage risk of the feeder, crusher and screening machine is evaluated respectively. The instantaneous load abnormality risk, predicted trend risk intensity, comprehensive blockage risk and blockage risk level of each equipment are integrated to obtain comprehensive information on blockage risk assessment.

[0089] By performing the above operations, this solution addresses the problem in the existing adaptive control process of sand and gravel crushing production. This is because the equipment load is significantly affected by feed fluctuations and changes in material properties, and material blockage often exhibits a gradual development characteristic. Existing methods often rely on current or vibration thresholds at a single moment for judgment, making it difficult to distinguish between short-term operating fluctuations and the actual evolution of material blockage. This leads to delayed identification of blockage risks, frequent false alarms, and consequently, control adjustments often occur after blockage has already formed or is close to forming. This solution creatively adopts a sand and gravel crushing blockage risk assessment method based on hierarchical fusion of spatiotemporal features. It achieves adaptive identification of abnormal equipment load under conditions of sudden feed changes and operating condition switching, and on this basis, it provides an early assessment of the short-term evolution trend of blockage risks, thus providing a forward-looking risk constraint basis for subsequent production control optimization.

[0090] Example 4, see Figure 1 and Figure 4 This embodiment is based on the above embodiment. In step S3, the production control optimization is used to generate a collaborative control strategy that takes into account production efficiency, product quality, energy consumption, and operational safety under the constraint of material blockage risk. Specifically, based on the multi-dimensional production dataset and the comprehensive information of material blockage risk assessment, an adaptive production control strategy package is obtained by adopting a risk-aware multi-agent collaborative game production control optimization method, including the following steps:

[0091] Step S31: Multi-agent construction, used to construct a collaborative decision-making environment for the sand and gravel crushing production process under the condition of material blockage risk perception. Specifically, by constructing the feeder, crusher and screening machine in the sand and gravel crushing production line as independent agents, and constructing agent state vectors for each agent, the multi-agent collaborative control environment model is performed to obtain a multi-agent collaborative decision-making environment model that incorporates material blockage risk information.

[0092] The agent state vector includes the equipment operating status, the comprehensive material blockage risk corresponding to each equipment, and the system-level comprehensive material blockage risk.

[0093] The system-level comprehensive material blockage risk is specifically obtained by weighting and fusing the comprehensive material blockage risks corresponding to each piece of equipment. It is used to characterize the overall material blockage risk level of the current sand and gravel crushing production system. The calculation formula is as follows:

[0094] ;

[0095] In the formula, R final (t) represents the system-level comprehensive material blockage risk at time step t, j is the common index of equipment and agents, J is the number of equipment, which is equal to the number of agents, and the equipment includes feeders, crushers, and screening machines. R is the risk contribution weight of the j-th device. j (t) represents the overall risk of material blockage for the j-th device at time step t;

[0096] Step S32: Control action space constraint, which is used to actively suppress the control behavior of each device when the risk of material blockage increases. Specifically, by setting corresponding control action variables for each agent and performing risk modulation on the control action space of each agent based on the comprehensive information of material blockage risk assessment, the control action range is adaptively shrunk to obtain a risk constraint control action space that matches the level of material blockage risk.

[0097] The formula for calculating the risk constraint control action space is as follows:

[0098] ;

[0099] In the formula, Let j be the risk constraint control action space of the j-th agent, where j is the agent index. It is the minimum permissible range of control action. It is the maximum permissible range of motion for controlling the movement;

[0100] The formula for calculating the maximum permissible control amplitude is as follows:

[0101] ;

[0102] In the formula, It is the rated maximum control amplitude of the j-th agent. It is the risk inhibition coefficient;

[0103] Step S33: Risk utility construction, used to achieve adaptive risk trade-off between overall system operating efficiency and equipment operating safety. Specifically, it involves constructing a global objective function and dynamically adjusting the objective weight based on the system-level comprehensive material blockage risk, while constructing local utility functions for each agent to obtain a game utility model.

[0104] The global objective function specifically uses the output, energy consumption, product particle size deviation, and system-level comprehensive blockage risk of the sand and gravel crushing production system as indicators, and the calculation formula is as follows:

[0105] ;

[0106] In the formula, G is the global objective function, w1 is the output target weight, Q is the output of the sand and gravel crushing production system, w2 is the energy consumption target weight, E is the energy consumption of the sand and gravel crushing production system, w3 is the particle size deviation target weight, D is the product particle size deviation, used to characterize the degree of deviation between the product particle size and the target particle size, w4 is the blockage risk target weight, and R... final It is a system-level comprehensive risk of material blockage;

[0107] The formula for calculating the dynamically adjusted target weight is as follows:

[0108] ;

[0109] ;

[0110] In the formula, This is the initial value for the target weight of material blockage risk. It is a weighting adjustment coefficient. This is the initial value for the production target weight;

[0111] The local utility function includes the global objective function and the device constraint penalty term, and the calculation formula is as follows:

[0112] ;

[0113] In the formula, U j It is the local utility function of the j-th agent. It is the safety weight of the j-th agent, Pen j It is the penalty for the j-th agent violating the risk constraint;

[0114] Step S34: Hierarchical game solution, used to generate a coordinated multi-device collaborative control strategy under risk constraints. Specifically, based on the preset initial control strategy, the operation results of the sand and gravel crushing production system under different combinations of control strategies are predicted by a process-level digital twin prediction model to obtain the system operation prediction output. Based on the system operation prediction output, the control actions of each agent are iteratively optimized using a hierarchical game update mechanism to obtain the optimal set of multi-agent collaborative control strategies.

[0115] The process-level digital twin prediction model is specifically implemented using a lightweight process modeling approach that combines historical operating data of the sand and gravel crushing production system with mechanistic constraints. It is used to characterize the mapping relationship between control strategies and the output, energy consumption, product particle size deviation, and system-level comprehensive material blockage risk of the sand and gravel crushing production system.

[0116] The system operation prediction output includes system output prediction, system energy consumption prediction, product particle size deviation prediction, and system-level comprehensive material blockage risk prediction.

[0117] The hierarchical game update mechanism involves updating the control actions of each agent in a predetermined order, based on their respective local utility functions, while keeping the control actions of other agents unchanged. This process gradually approaches the stable state of the overall system utility. The calculation formula is as follows:

[0118] ;

[0119] In the formula, It represents the updated control action of the j-th agent, where d is the iteration index. It is the control action of the j-th agent in the d-th iteration. It is an update of the compensation coefficient. It is the gradient of the control action of the j-th agent;

[0120] Preferably, the preset sequence is specifically based on the process flow sequence of the feeder, crusher and screening machine, and is dynamically adjusted in combination with the real-time blockage risk level of each equipment. When the overall blockage risk of downstream equipment is high, its corresponding control strategy is updated first, and then the optimization results are passed on to upstream equipment in a stepwise manner.

[0121] The calculation formula for the optimal cooperative control strategy set of the multi-agent system is as follows:

[0122] ;

[0123] In the formula, It is a set of optimal cooperative control strategies for multiple agents. This is the optimal feeding speed for the feeder. These are the optimal discharge port parameters for the crusher. This is the optimal speed for the crusher. This is the optimal vibration frequency for the screening machine. It is the optimal amplitude for the screening machine;

[0124] Step S35: Strategy package generation, specifically, by integrating the optimal collaborative control strategy set of multi-agents, an adaptive production control strategy package containing the optimal control instructions of each device is generated, and the corresponding applicable material blockage risk range, expected operating effect evaluation information, and risk over-limit retreat conditions are associated in the adaptive production control strategy package.

[0125] By performing the above operations, this solution addresses the technical problems in the existing adaptive control process of sand and gravel crushing production. These problems include the independent adjustment of key equipment according to their respective operating conditions, with overall control primarily based on single-machine optimization, making it difficult to form a collaborative optimization mechanism for the entire production line. Furthermore, when the risk of material blockage increases, the production system struggles to effectively balance output, energy consumption, and operational safety, easily leading to incoordination between upstream feeding and downstream processing, frequent speed limits or shutdowns of local equipment, and significant fluctuations in overall production efficiency. This solution creatively adopts a risk-aware multi-agent collaborative game-theoretic production control optimization method. This method achieves coordinated adjustment of the control behavior of multiple equipment such as feeding, crushing, and screening under the constraint of material blockage risk. It enables the control strategies of each equipment to dynamically converge to a safe, stable, and efficient operating state as the risk changes, thereby reducing the probability of material blockage and improving overall operational efficiency while ensuring production continuity.

[0126] Example 5, see Figure 1 This embodiment is based on the above embodiment. In step S4, the adaptive control execution specifically involves adaptively controlling the sand and gravel crushing production process according to the adaptive production control strategy package, continuously monitoring the changes in key operating indicators and the comprehensive risk of material blockage during the control execution process, and achieving adaptive stable operation and material blockage risk suppression of the sand and gravel crushing production process by adopting strategy effect feedback and risk trigger backoff control operations, and generating an adaptive control execution report.

[0127] The adaptive control execution report includes, but is not limited to, the actual execution control parameters of each device, the real-time comprehensive risk change curve of material blockage, the execution effect evaluation index, and the records of anomalies and rollbacks;

[0128] The performance evaluation indicators include output achievement rate, risk reduction rate, and energy consumption change.

[0129] Example 6, see Figure 2 Based on the above embodiments, this embodiment provides an adaptive control system for sand and gravel crushing production process, including a data acquisition module, a risk assessment module, a control optimization module, and a control execution module.

[0130] The data acquisition module is used for production data acquisition. Through production data acquisition, a multidimensional production dataset is obtained, and the multidimensional production dataset is sent to the risk assessment module and the control optimization module.

[0131] The risk assessment module is used for material blockage risk assessment. Through the material blockage risk assessment, comprehensive material blockage risk assessment information is obtained, and the comprehensive material blockage risk assessment information is sent to the control optimization module.

[0132] The control optimization module is used for production control optimization. Through production control optimization, an adaptive production control strategy package is obtained, and the adaptive production control strategy package is sent to the control execution module.

[0133] The control execution module is used for adaptive control execution and to obtain an adaptive control execution report.

[0134] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0135] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention.

[0136] The present invention and its embodiments have been described above. This description is not restrictive, and the accompanying drawings are only one embodiment of the present invention; the actual structure is not limited thereto. In conclusion, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the invention, such designs should fall within the protection scope of the present invention.

Claims

1. An adaptive control method for a sand and gravel crushing production process, characterized in that: The method includes the following steps: Step S1: Production data collection to obtain a multidimensional production dataset; Step S2: Blockage risk assessment. Specifically, based on the multidimensional production dataset, a blockage risk assessment method for sand and gravel crushing based on spatiotemporal feature hierarchical fusion is adopted to obtain comprehensive information on blockage risk assessment. This includes the following steps: Step S21 Load deviation construction, Step S22 State trend prediction, Step S23 Hierarchical risk fusion, Step S24 Risk level determination, and Step S25 Comprehensive information generation. In step S21, the load deviation construction specifically involves performing dynamic baseline modeling on the main motor current signal and the equipment vibration signal to obtain the dynamic mean and dynamic dispersion of the corresponding signals, and standardizing the load signal at the current time step based on the dynamic mean and dynamic dispersion to construct a multi-source load deviation vector and calculate the instantaneous load anomaly intensity. The dynamic baseline modeling specifically involves statistically modeling the main motor current signal and the equipment vibration signal using an exponential sliding update method, respectively, to obtain the dynamic mean and dynamic dispersion that are adaptively adjusted with changes in operating conditions in real time, so as to characterize the normal load baseline level under the current operating state. In step S22, the state trend prediction is used to assess the evolution trend of material blockage risk in the near future. Specifically, it involves constructing a multivariate historical time series as input data, predicting the main motor load and equipment vibration state through a long short-term memory network, and calculating the predicted trend risk intensity based on the degree of deviation between the prediction results and the current dynamic baseline. The multivariate historical time series includes feed rate, main motor current, equipment vibration, equipment temperature, and discharge port pressure; The prediction results include a main motor current prediction sequence and a device vibration prediction sequence; In step S23, the hierarchical risk fusion specifically involves mapping the instantaneous load anomaly intensity to instantaneous load anomaly risk through nonlinear mapping, introducing a predicted trend risk gating factor to dynamically adjust the predicted risk weight, and combining it with historical risk accumulation terms to perform weighted fusion of multi-level risk information to obtain the comprehensive risk of material blockage. The multi-level risk information includes instantaneous load anomaly risk, predicted trend risk intensity, and historical risk accumulation. Step S3: Production control optimization, used to generate a collaborative control strategy that takes into account production efficiency, product quality, energy consumption and operational safety under the constraint of material blockage risk. Specifically, based on the multi-dimensional production dataset and comprehensive information on material blockage risk assessment, a risk-aware multi-agent collaborative game production control optimization method is adopted to obtain an adaptive production control strategy package, including the following steps: Step S31 Multi-agent construction, Step S32 Control action space constraint, Step S33 Risk utility construction, Step S34 Hierarchical game solution and Step S35 Strategy package generation; In step S31, the multi-agent construction is used to construct a collaborative decision-making environment for the sand and gravel crushing production process under the condition of material blockage risk perception. Specifically, it involves constructing the feeder, crusher and screening machine in the sand and gravel crushing production line as independent agents, constructing agent state vectors for each agent, and modeling the multi-agent collaborative control environment to obtain the multi-agent collaborative decision-making environment model. Step S4: Adaptive control execution.

2. The adaptive control method for sand and gravel crushing production process according to claim 1, characterized in that: In step S24, the risk level determination is used to classify the risk of material blockage. Specifically, it maps the comprehensive risk of material blockage to a preset risk range through nonlinear mapping to obtain the risk level of material blockage. In step S25, the comprehensive information generation specifically involves performing steps S21 to S24 to assess the material blockage risk of the feeder, crusher, and screening machine, and integrating the instantaneous load anomaly risk, predicted trend risk intensity, comprehensive material blockage risk, and material blockage risk level of each device to obtain comprehensive material blockage risk assessment information.

3. The adaptive control method for sand and gravel crushing production process according to claim 2, characterized in that: In step S31, the agent state vector includes the device operating status, the overall material blockage risk corresponding to each device, and the overall material blockage risk at the system level; The system-level comprehensive material blockage risk is specifically obtained by weighted fusion of the comprehensive material blockage risks corresponding to each device; In step S32, the control action space constraint is used to actively suppress the control behavior of each device when the risk of material blockage increases. Specifically, it is achieved by setting corresponding control action variables for each intelligent agent and performing risk modulation on the control action space of each intelligent agent based on the comprehensive information of material blockage risk assessment, thereby adaptively shrinking the control action range to obtain the risk-constrained control action space. In step S33, the risk utility construction is used to achieve an adaptive risk trade-off between the overall system operating efficiency and equipment operating safety. Specifically, it involves constructing a global objective function and dynamically adjusting the objective weight based on the system-level comprehensive material blockage risk, while constructing local utility functions for each agent to obtain a game utility model. The global objective function specifically uses the output, energy consumption, product particle size deviation, and system-level comprehensive material blockage risk of the sand and gravel crushing production system as indicators.

4. The adaptive control method for sand and gravel crushing production process according to claim 3, characterized in that: In step S34, the hierarchical game solution specifically involves predicting the operation results of the sand and gravel crushing production system under different combinations of control strategies based on a preset initial control strategy using a process-level digital twin prediction model, obtaining the system operation prediction output, and iteratively optimizing the control actions of each agent using a hierarchical game update mechanism based on the system operation prediction output, thereby obtaining the optimal collaborative control strategy set for multiple agents. The system operation prediction output includes system output prediction, system energy consumption prediction, product particle size deviation prediction, and system-level comprehensive material blockage risk prediction. The hierarchical game update mechanism is specifically as follows: while keeping the control actions of other agents unchanged, each agent updates its control actions one by one according to the corresponding local utility function in a preset order, so as to gradually approach the stable state of the overall system utility. In step S35, the strategy package is generated by integrating the optimal collaborative control strategy set of multi-agents to generate an adaptive production control strategy package containing the optimal control instructions of each device, and by associating the corresponding applicable material blockage risk range, expected operating effect evaluation information, and risk over-limit retreat conditions in the adaptive production control strategy package.

5. The adaptive control method for sand and gravel crushing production process according to claim 4, characterized in that: In step S4, the adaptive control execution specifically involves adaptively controlling the sand and gravel crushing production process based on the adaptive production control strategy package, continuously monitoring key operating indicators and changes in material blockage risk during the control execution process, and achieving adaptive stable operation and material blockage risk suppression of the sand and gravel crushing production process by adopting strategy effect feedback and risk-triggered rollback control operations, thereby generating an adaptive control execution report.

6. The adaptive control method for a sand and gravel crushing production process according to claim 5, characterized in that: In step S1, the production data acquisition specifically involves acquiring raw signal data through current sensors, vibration sensors, temperature sensors, feed rate detection devices, and discharge port pressure detection devices installed on the feeder, crusher, and screening machine, and performing time synchronization, multi-source data cleaning, and feature standardization processing to obtain a multi-dimensional production dataset.

7. An adaptive control system for a sand and gravel crushing production process, used to implement the adaptive control method for a sand and gravel crushing production process as described in any one of claims 1-6, characterized in that: It includes a data acquisition module, a risk assessment module, a control optimization module, and a control execution module; The data acquisition module is used for production data acquisition. Through production data acquisition, a multidimensional production dataset is obtained, and the multidimensional production dataset is sent to the risk assessment module and the control optimization module. The risk assessment module is used for material blockage risk assessment. Through the material blockage risk assessment, comprehensive material blockage risk assessment information is obtained, and the comprehensive material blockage risk assessment information is sent to the control optimization module. The control optimization module is used for production control optimization. Through production control optimization, an adaptive production control strategy package is obtained, and the adaptive production control strategy package is sent to the control execution module. The control execution module is used for adaptive control execution and to obtain an adaptive control execution report.

Images