Intelligent control method and system of an electric dust precipitator

Abstract

The application discloses an intelligent regulation and control method and system of an electric dust collector, and relates to the technical field of intelligent regulation and control.The method comprises the following steps: collecting data through a sensor, constructing a mathematical model of an electric dust removal process, and simplifying a state space model by using a deep learning model; estimating current state parameters of the state space model based on sensor data, recursively calculating future outputs in a given prediction time domain and control time domain in combination with the deep learning model; setting an adaptive weighted quadratic objective function of tracking error and control increment, and outputting safety constraints; calculating an optimal control parameter sequence by using a genetic algorithm in combination with the objective and the constraints, executing first-order control, and rolling updating the control parameters.The method significantly improves the control accuracy and response stability of the electric dust collector, effectively reduces the energy consumption of the system, and improves the dust removal efficiency.Meanwhile, the adaptability of the system is further enhanced, so that the system can maintain efficient operation under different working conditions.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology, and in particular to an intelligent control method and system for electrostatic precipitators. Background Technology

[0002] Electrostatic precipitators (ESPs), as an important air purification method, are widely used in industrial production processes, especially in the power, metallurgy, and chemical industries, to remove dust particles from exhaust gases. Traditional ESPs use a high-voltage electric field to charge dust particles, which are then adsorbed by the collecting electrodes, thus purifying the air. However, with the acceleration of industrialization and increasingly stringent environmental protection standards, ESP technology faces growing challenges in improving dust removal efficiency, reducing energy consumption, and ensuring equipment stability. Currently, most traditional ESP control systems rely on experience and timed control, which struggles to cope with dynamic changes under different operating conditions and achieves precise control and optimization. To address this issue, recent research has focused on optimizing ESP control systems through intelligent methods, such as fuzzy control, expert systems, and neural networks, to improve control accuracy and adaptability.

[0003] However, existing electrostatic precipitator (ESP) control technologies still have shortcomings. Traditional control systems often cannot respond in real time to rapid changes in the ESP process, such as the impact of airflow fluctuations and changes in electric field strength on dust removal efficiency. This results in unstable performance of the control system under different load conditions. Although deep learning-based models have achieved significant results in some areas, there is a lack of systematic research on modeling electric fields and dust states in the intelligent control of ESPs. Consequently, many existing deep learning models do not fully consider the complex physical environment and variable operating conditions of ESPs. Therefore, existing technologies often cannot effectively predict and adjust in dynamic environments, have poor adaptability of control strategies, and struggle to guarantee system stability and efficiency. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides an intelligent control method for electrostatic precipitators to solve the problem that traditional technologies lack the ability to respond quickly to dynamically changing environments and cannot accurately control factors such as electric field strength and airflow changes.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides an intelligent control method for an electrostatic precipitator, which includes collecting data through sensors, constructing a mathematical model of the electrostatic precipitator process, and simplifying it using a deep learning model to obtain a state-space model.

[0008] The current state parameters of the state-space model are estimated based on sensor data, and the future output is recursively inferred by combining a deep learning model within the given prediction and control time domains.

[0009] Define an adaptive weighted quadratic objective function for tracking error and control increment, and output safety constraints;

[0010] The optimal control parameter sequence is calculated by combining the objective and constraints using a genetic algorithm, and the first-order control is executed while the control parameters are updated on a rolling basis.

[0011] As a preferred embodiment of the intelligent control method for the electrostatic precipitator described in this invention, the simplified mathematical model for constructing the electrostatic precipitator process includes constructing a mathematical model of the electric field and dust state within the electrostatic precipitator. The specific modeling process includes:

[0012] Assuming the electric field strength of the electrostatic precipitator is a function of space, the electric field distribution is described using the Poisson equation based on the electrode shape and voltage. By applying a voltage between the electrodes of the electrostatic precipitator, the relationship between current density and electric field strength is calculated.

[0013] The dust state is composed of dust movement and dust concentration; dust particles are affected by gravity, airflow drag force and electric field force, which constitute the motion equation of dust particles; within the working space of the electrostatic precipitator, the change of dust particle concentration is affected by the combined effect of airflow and electric field, and the change of dust concentration is described by the continuous medium equation.

[0014] As a preferred embodiment of the intelligent control method for the electrostatic precipitator described in this invention, the step of simplifying using a deep learning model includes: pre-training a deep neural network, inputting electric field model parameters and dust state model parameters within the electrostatic precipitator; modeling the relationship between static input and output, and learning the nonlinear mapping between input and output through a multi-layer fully connected network;

[0015] By extracting nonlinear relationships from the input data, the output obtained by training the deep learning model is used as the state parameters of the state-space model. The mathematical model is simplified and state transitions are performed to obtain the state-space model.

[0016] As a preferred embodiment of the intelligent control method for the electrostatic precipitator of the present invention, the step of combining the deep learning model to predict the future output includes fusing the Kalman filter estimate with the prediction value of the deep learning model to output the future state parameters.

[0017] Using data acquired from sensors, the Kalman filter method is used to estimate the current state parameters in the mathematical model, yielding the estimation results. Among them, the prior estimate is obtained by the deep learning model based on historical data;

[0018] Set the current state parameters and control input parameter u t Input a deep learning model, and through the forward propagation process, output the predicted state parameters x for the next time step. t+1 The prediction time domain and control time domain are preset to determine the prediction range of future outputs; the future prediction time domain is set to start from the current time t and predict the state for the next K time steps; during the prediction process, the state parameters x for the next K time steps are recursively calculated using a deep learning model. t+k ;

[0019] The state estimation results of the Kalman filter The prediction result x of the deep learning model t+k Perform weighted fusion to obtain the future state result.

[0020] As a preferred embodiment of the intelligent control method for the electrostatic precipitator described in this invention, the adaptive weighted quadratic objective function includes the input state vector x of the current system. t Define the reference state vector as x ref,t Indicates the expected target value;

[0021] Define tracking error e t , representing the state-space parameter x t With the expected target value x ref,t The difference between them; control increment Δu t It is the control input u at the current moment. t and the control input u from the previous moment t-1 The amount of change between;

[0022] The objective function is constructed by including a weighted quadratic term for the tracking error and a weighted quadratic term for the control increment; the weights of the tracking error term are determined by the weighting matrix Q. e The goal of control is to minimize tracking error;

[0023] The weighted quadratic term of the control increment includes the weights of the control increment term determined by the weighting matrix R. u Control refers to the degree of penalty imposed on the increment of the control input; the goal is to minimize the change in the control increment.

[0024] Add control input u t The constraints make u t Between the effective range of the control input; for the control increment Δu t The constraint is always applied and must be less than the maximum permissible range of change in the control input.

[0025] As a preferred embodiment of the intelligent control method for the electrostatic precipitator described in this invention, the step of using a genetic algorithm to calculate the optimal control parameter sequence by combining the objective and constraints includes initializing a population of the genetic algorithm, where each individual in the population represents a control sequence, and each control input u t It is a decision variable in the control time domain; generate the initial population, where each control input parameter u t Randomly generated within the control input range;

[0026] Define a fitness function and use an objective function to evaluate the criteria for each individual's performance; for each individual's sequence, u t The evaluation is performed using a weighted quadratic objective function with additional constraints.

[0027] Using the roulette wheel selection method, individuals with high fitness are selected from the current population as parents for crossover and mutation operations; within the constraints, the u in the sequence... t Fitness weights decrease sequentially.

[0028] Perform crossover on the selected parent individuals to generate new offspring individuals; select two parent control sequences and perform crossover to generate a new control sequence.

[0029] For the input u in the new control sequence t Perform random perturbations, and ensure that the mutated control input still meets the control input range constraints and incremental constraints, and iterate to obtain the optimal control input parameter sequence.

[0030] As a preferred embodiment of the intelligent control method for the electrostatic precipitator described in this invention, the rolling update of control parameters includes, after obtaining the optimal control sequence, performing first-order control by using the first control input parameter in the control sequence. To control the operation of the electrostatic precipitator, input the data into the electrostatic precipitator's control system, and adjust the system status;

[0031] After the first-order control is executed, the sensors update the current system state based on the feedback values. After the control operation is executed, the state-space parameters of the electrostatic precipitator change; the system state is then updated based on the data provided by the sensors and the current control strategy. The method of using Kalman filtering algorithm and deep learning model prediction fusion is corrected based on actual observations and predictions;

[0032] Let λ u (t) represents the weighting coefficient of the control increment in the objective function, which is adjusted according to the rate of change of the control input; when the change in the control input approaches the constraint range, the weight is increased to smooth the control process; λ e(t) represents the tracking error weight in the objective function. As the control error increases, λ increases. e Adjust the (t) value;

[0033] Based on the latest state estimate Based on the reference trajectory, recalculate the objective function and perform rolling updates of the optimal control sequence. Recalculate the optimal solution of the objective function and update the control strategy; then apply the new system state. As input, the objective function is solved again to generate a new control input sequence.

[0034] Secondly, the present invention provides an intelligent control system for an electrostatic precipitator, comprising an electrostatic precipitator modeling module, a deep learning prediction module, and an adaptive optimization module;

[0035] The electrostatic precipitator modeling module is used to construct a mathematical model of the electric field and dust state inside the electrostatic precipitator.

[0036] The deep learning prediction module is used to simplify complex mathematical models and estimate future outputs.

[0037] The adaptive optimization module uses a genetic algorithm to calculate the optimal control sequence and updates the system state based on sensor feedback data.

[0038] Thirdly, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements any step of the intelligent control method for an electrostatic precipitator as described in the first aspect of the present invention.

[0039] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the intelligent control method for an electrostatic precipitator as described in the first aspect of the present invention.

[0040] The beneficial effects of this invention are as follows: By combining deep learning and Kalman filtering, the control accuracy and response stability of the electrostatic precipitator are significantly improved. Compared with the frequent adjustments and large changes that occur in traditional control methods, this invention effectively reduces system energy consumption and improves dust removal efficiency by smoothing control increments and optimizing the control sequence. At the same time, the introduction of a genetic algorithm further enhances the system's adaptability, enabling it to maintain high-efficiency operation under different working conditions. Attached Figure Description

[0041] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 This is a flowchart of an intelligent control method for electrostatic precipitators. Detailed Implementation

[0043] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0044] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0045] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0046] Reference Figure 1 As one embodiment of the present invention, this embodiment provides an intelligent control method for an electrostatic precipitator, comprising the following steps:

[0047] S1: Collect data through sensors, construct a mathematical model of the electrostatic precipitator process, and simplify it using a deep learning model to obtain a state-space model.

[0048] Furthermore, the simplified mathematical model for constructing the electrostatic precipitator process includes constructing a mathematical model of the electric field and dust state within the electrostatic precipitator. The specific modeling process includes:

[0049] Assuming the electric field strength of the electrostatic precipitator is a function of space, the electric field distribution is described using the Poisson equation based on the electrode shape and voltage. By applying a voltage between the electrodes of the electrostatic precipitator, the relationship between current density and electric field strength is calculated.

[0050] The dust state is composed of dust movement and dust concentration; dust particles are affected by gravity, airflow drag force and electric field force, which constitute the motion equation of dust particles; within the working space of the electrostatic precipitator, the change of dust particle concentration is affected by the combined effect of airflow and electric field, and the change of dust concentration is described by the continuous medium equation.

[0051] The simplification using deep learning models includes: pre-training a deep neural network, inputting electric field model parameters and dust state model parameters into the electrostatic precipitator; modeling the relationship between static input and output, and learning the nonlinear mapping between input and output through a multi-layer fully connected network.

[0052] By extracting nonlinear relationships from the input data, the output obtained by training the deep learning model is used as the state parameters of the state-space model. The mathematical model is simplified and state transitions are performed to obtain the state-space model.

[0053] The mathematical model of the electrostatic precipitator process can be divided into two parts: the electric field model and the dust state model. First, these two models are established to describe the electric field distribution and the dust motion state, respectively. The electric field model is one of the most important factors in an electrostatic precipitator. It is assumed that the electric field strength E(x,y,z) is a function in space, depending on the shape of the electrodes and the voltage. The Poisson equation is used to describe the electric field distribution:

[0054]

[0055] Where ρ(x,y,z) is the charge density, representing the charge distribution in the electric field. ε is the permittivity, representing the conductivity of the medium. To relate the electric field to the current density, Ohm's law is used to obtain the relationship between the current density J(x,y,z) and the electric field strength E(x,y,z):

[0056] J(x,y,z)=σ(x,y,z)E(x,y,z)

[0057] Where σ(x,y,z) is the electrical conductivity, which represents the material's ability to conduct electric current.

[0058] Dust state model: The motion and concentration of dust are influenced by multiple factors: gravity, airflow drag force, and electric field force. Assuming the mass of the dust particle is m, its equation of motion can be expressed as:

[0059]

[0060] Wherein: F g =mg is the gravitational force. F d =αv is the airflow drag force, where α is the drag coefficient and v is the particle velocity. F e =qE is the electric field force, q is the charge of the dust particle, and E is the electric field strength. The dust particle concentration C(x,y,z,t) varies with time in the working space of the electrostatic precipitator, influenced by airflow and the electric field. The continuous medium equation can describe the change in dust concentration:

[0061]

[0062] in, It is the diffusion flow of dust, representing the movement and concentration changes of dust particles.

[0063] The simplified state-space model uses a deep learning model to simplify the complex mathematical model described above.

[0064] The pre-training of the deep learning model uses a deep neural network (DNN) to learn the nonlinear relationship between the electric field model and the dust state model. The inputs are parameters of the electric field model and the dust state model, such as voltage, airflow velocity, and dust particle density.

[0065] The mapping relationship between inputs and outputs is learned through multi-layer fully connected networks (MLPs). The structure of a neural network typically consists of fully connected layers composed of multiple hidden layers, each receiving input... and output There is a non-linear mapping between them:

[0066]

[0067] in, It is a weight matrix. is the bias term, and f(·) is the activation function. The network is trained using the backpropagation algorithm, and the weights and bias terms are adjusted to minimize the network's prediction error.

[0068] Once the state-space model is trained, the deep learning model can infer the state parameters x of the state-space model based on the input data. k The state-space model takes the following form:

[0069] x k =f θ (x k-1 ,u k )+w k

[0070] Where: x k This is the state parameter at the current moment. k It is a control input. f θ It is a nonlinear state transition function obtained from a deep learning model. k It is process noise.

[0071] It should be noted that the training data and input data include parameters of the electric field model (such as voltage, current density, and airflow velocity) and parameters of the dust state model (such as particle concentration and particle velocity), which are collected in real time by sensors. By inputting this data, the deep learning network can learn the nonlinear mapping relationship between input and output, thereby modeling the complex dynamic behavior of the electrostatic precipitator process. Deep learning models can handle complex nonlinear relationships between input data, while traditional physical modeling methods often assume a linear relationship. This makes deep learning models more flexible and efficient than traditional methods when dealing with complex dynamic systems (such as the interaction between airflow, dust, and electric field in the electrostatic precipitator process). In this way, deep learning models simplify the nonlinear complexity of the electrostatic precipitator process, making it easier to use for subsequent control strategy optimization.

[0072] S2: Estimate the current state parameters of the state-space model based on sensor data, and recursively predict the future output by combining the deep learning model within the given prediction and control time domains.

[0073] Furthermore, the step of combining the deep learning model to predict the future output includes fusing the Kalman filter estimate with the prediction of the deep learning model to output the future state parameters.

[0074] Using data acquired from sensors, the Kalman filter method is used to estimate the current state parameters in the mathematical model, yielding the estimation results. The prior estimates are obtained by the deep learning model based on historical data.

[0075] Set the current state parameters and control input parameter u t Input the deep learning model, and through the forward propagation process, output the predicted state parameters for the next time step. x t +1 The prediction time domain and control time domain are preset to determine the prediction range of future outputs; the future prediction time domain is set to start from the current time t and predict the state for the next K time steps; during the prediction process, the state parameters x for the next K time steps are recursively calculated using a deep learning model. t+k .

[0076] The state estimation results of the Kalman filter The prediction result x of the deep learning model t+k Perform weighted fusion to obtain the future state result.

[0077] In control systems, it is necessary to estimate the current state of the system and predict its future state. By combining Kalman filtering and deep learning models, more accurate state estimation and prediction can be achieved.

[0078] Kalman Filtering State Estimation: The Kalman filtering method is used to estimate the state of the current system. Based on sensor observations y k and prior state estimation

[0079]

[0080] Where: A is the state transition matrix, representing the state change from time k-1 to time k. B is the control input matrix, representing the influence of the control input on the state change. K k is the Kalman gain, which controls the degree of fusion between the estimated and observed values. C is the observation matrix, representing the mapping from states to observations.

[0081] Deep learning models predict future states based on historical data:

[0082]

[0083] The Kalman filter estimates are fused with the predictions from the deep learning model:

[0084]

[0085] in: This is the state estimated by the Kalman filter. This represents the state predicted by the deep learning model. α is the fusion coefficient, which determines the weights of the Kalman filter and the deep learning model.

[0086] It should be noted that the state parameter x output by the deep learning model is used. t This approach simplifies the complex mathematical model of electrostatic precipitators into a state-space model. Traditional mathematical models contain multiple physical equations (such as the Poisson equation, electric field distribution, and dust motion equations), which require complex numerical calculations. Deep learning models can automatically extract nonlinear relationships from input data and simplify them into a state-space model. The dynamic updates of this model are then accomplished through the predictive capabilities of deep learning. This simplification is highly innovative, significantly reducing reliance on complex physical formulas while improving the system's adaptability to external changes (such as uneven electric fields and airflow variations).

[0087] S3: Set an adaptive weighted quadratic objective function for tracking error and control increment, and output safety constraints.

[0088] Furthermore, the adaptive weighted quadratic objective function includes the input current system state vector x. t Define the reference state vector as x ref,t This represents the expected target value.

[0089] Define tracking error e t , representing the state-space parameter x t With the expected target value x ref,t The difference between them; control increment Δu t It is the control input u at the current moment. t and the control input u from the previous moment t-1 The amount of change between.

[0090] The objective function is constructed by including a weighted quadratic term for the tracking error and a weighted quadratic term for the control increment; the weights of the tracking error term are determined by the weighting matrix Q. e The goal of control is to minimize tracking error.

[0091] The weighted quadratic term of the control increment includes the weights of the control increment term determined by the weighting matrix R. u Control refers to the degree of penalty imposed on the increment of the control input; the goal is to minimize the change in the control increment.

[0092] Add control input u t The constraints make u t Between the effective range of the control input; for the control increment Δu t The constraint is always applied and must be less than the maximum permissible range of change in the control input.

[0093] The objective function is designed to take into account a weighted quadratic term of tracking error and control increment. By adjusting the weighting coefficients, a balance can be struck between tracking accuracy and control stability.

[0094] The objective function J consists of two parts: a weighted quadratic term of the tracking error.

[0095] J e =w e ||x k -x ref,k || 2

[0096] Where, x ref,k It is the desired state, w e These are the weighting coefficients for the tracking error. The weighted quadratic term for the control increment is expressed by the formula:

[0097] J u =w u ||Δu k || 2

[0098] Where, Δu k =u k -u k-1 It controls the increment, w u These are the weighting coefficients that control the increment. The final objective function is:

[0099] J = J e +J u

[0100] By minimizing the objective function, the control system can find the optimal control input sequence, thereby achieving precise control of the electrostatic precipitator.

[0101] It should be noted that the fusion coefficient α is dynamically adjusted, rather than statically set. This innovation ensures that the system can adaptively adjust the weights of the deep learning model and the Kalman filter model based on different operating conditions and error profiles. For example, when noise is high or sensor errors are significant, more reliance may be placed on the Kalman filter; while when the system is more stable and the deep learning model can provide accurate predictions, the weights of the deep learning model can be increased.

[0102] S4: Use a genetic algorithm to calculate the optimal control parameter sequence by combining the objective and constraints, execute the first-order control and update the control parameters in a rolling manner.

[0103] Furthermore, the step of using a genetic algorithm to calculate the optimal control parameter sequence by combining the objective and constraints includes initializing the genetic algorithm population, where each individual in the population represents a control sequence, and each control input u t It is a decision variable in the control time domain; generate the initial population, where each control input parameter u t Randomly generated within the control input range.

[0104] Each individual in the population represents a control sequence, where each control input is a decision variable in the control time domain. The specific process is as follows: Population definition: Each individual in the population represents a control sequence u. k =[u1,u2,...,u N ], where u k It is the control input sequence within the control time domain, where N is the length of the control time domain. The initial control input is randomly generated: within the effective range of the control input [u...]. min ,u max Each control input is randomly generated within [the specified range]. That is, each control input u... k satisfy:

[0105] u k ∈[u min ,u max ]

[0106] After the initial population is generated, the entire population is represented as a set of multiple control inputs.

[0107] The fitness function and objective function are defined as follows: The fitness function is the standard used in genetic algorithms to evaluate the quality of individuals. The fitness function is defined as the evaluation standard of the objective function, which includes a weighted quadratic objective function of tracking error and control increment. The specific steps are as follows: Objective function: The objective function consists of two parts: the tracking error part J... e :

[0108] J e =w e ||x k -x ref,k || 2

[0109] Where, x k This is the current system state, x ref,k It is the expected target value, w e This is the weighting coefficient for the tracking error. The control increment J... u :

[0110] J u =w u ||Δu k || 2

[0111] Where, Δu k =u k -u k-1 It controls the increment, w u These are the weighting coefficients that control the increment. The overall objective function J is the weighted sum of these two parts:

[0112] J = J e +J u

[0113] Constraints: When calculating fitness, constraints are added to the objective function. Constraints on the control increment and the effective range of the control input are incorporated into the weighted quadratic objective function. The constraints can be expressed as:

[0114] u min ≤u k ≤u max

[0115] And constraints on the change of control increment:

[0116] Δu k ≤Δu max

[0117] In each generation of a genetic algorithm, roulette wheel selection and crossover operations are used to select individuals with higher fitness as parents for crossover and mutation operations.

[0118] Roulette wheel selection: This method calculates the fitness F of each individual.i =-J i (Negative values, because we want to minimize the objective function), calculate the choice probability for each individual:

[0119]

[0120] Then, probability is used for selection, choosing individuals with high fitness as parents for crossover and mutation. Crossover operation: Select two parent individuals u. k1 and u k2 Two new control sequences u′ are generated through crossover operations. k1 and u′ k2 :

[0121] u′ k1 =αu k1 +(1-α)u k2

[0122] u′ k2 =(1-α)u k1 +αu k2

[0123] Where α∈[0,1] is the crossover coefficient, which controls how the new control sequence is generated.

[0124] To increase diversity and avoid local optima, the genetic algorithm also needs to perform a mutation operation. The mutation operation randomly perturbs the inputs in the newly generated control sequence. The specific steps are as follows: for each control input u in the new control sequence... k ', Perform random perturbation operation:

[0125] u k ′=u k ′+δ k

[0126] Where, δ k It is a random disturbance term, and δ k Satisfy the effective range constraint of the control input [u min ,u max [And constraints on control increments. The mutated control input u] k The control input range constraint and incremental constraint still need to be satisfied:

[0127] u min ≤u k ′≤u max

[0128] Δu k ′≤Δu max

[0129] After obtaining the optimal control sequence, the rolling update of control parameters executes the first-order control using the first control input parameter u1. Through the execution of the control input, the state-space parameters of the electrostatic precipitator change.

[0130] First-order control execution: Based on the optimized control sequence, the first control input u1 is used to control the electrostatic precipitator. Control input u1 is fed into the control system, and the operation of the electrostatic precipitator is adjusted to optimize its state.

[0131] After executing the first-order control, the sensor updates the current system state based on the feedback value. Assume the new system state is x. new,k Based on the results of the control operations, the system's state-space parameters changed. Kalman filtering and deep learning model fusion: State correction is performed using a fusion method of the Kalman filtering algorithm and a deep learning model. The Kalman filter is used to estimate the current state and then fused with the value predicted by the deep learning model. Predicted state. Recursively derived through deep learning models:

[0132]

[0133] The Kalman filter estimation result is then weighted and fused with the deep learning model's prediction result to obtain the updated state. The objective function is dynamically adjusted: the weights in the objective function are updated based on the new state estimate and the reference trajectory. Specifically, when the change in control increment approaches the constraint range, the weights need to be increased to ensure a smooth control process. The tracking error weights are dynamically adjusted as the control error increases to better track the target trajectory.

[0134] Define a fitness function and use an objective function to evaluate the criteria for each individual's performance; for each individual's sequence, u t The evaluation is performed using a weighted quadratic objective function with additional constraints.

[0135] Using the roulette wheel selection method, individuals with high fitness are selected from the current population as parents for crossover and mutation operations; within the constraints, the u in the sequence... t Fitness weights decrease sequentially.

[0136] Perform a crossover operation on the selected parent individuals to generate new offspring individuals; select two parent control sequences and perform a crossover operation to generate a new control sequence.

[0137] For the input u in the new control sequence t Perform random perturbations, and ensure that the mutated control input still meets the control input range constraints and incremental constraints, and iterate to obtain the optimal control input parameter sequence.

[0138] The rolling update of control parameters includes, after obtaining the optimal control sequence, performing first-order control and using the first control input parameter in the control sequence. It controls the operation of the electrostatic precipitator, inputs the data into the electrostatic precipitator's control system, and adjusts the system status.

[0139] After the first-order control is executed, the sensors update the current system state based on the feedback values. After the control operation is executed, the state-space parameters of the electrostatic precipitator change; the system state is then updated based on the data provided by the sensors and the current control strategy. The method of using Kalman filtering algorithm and deep learning model prediction fusion is used to correct the actual observations and predictions.

[0140] Subsequent parameters in the sequence are selected for adjustment. If they do not match the correction results, the objective function weights are modified to regenerate the future prediction sequence.

[0141] Let λ u (t) represents the weighting coefficient of the control increment in the objective function, which is adjusted according to the rate of change of the control input; when the change in the control input approaches the constraint range, the weight is increased to smooth the control process; λ e (t) represents the tracking error weight in the objective function. As the control error increases, λ increases. e Adjust the (t) value;

[0142] Based on the latest state estimate Based on the reference trajectory, recalculate the objective function and perform rolling updates of the optimal control sequence. Recalculate the optimal solution of the objective function and update the control strategy; then apply the new system state. As input, the objective function is solved again to generate a new control input sequence.

[0143] It should be noted that the weights in the objective function are updated based on the new state estimate and reference trajectory. Specifically, when the change in control increment approaches the constraint range, the weights need to be increased to ensure a smooth control process. The tracking error weights are dynamically adjusted as the control error increases to better track the target trajectory. Based on the latest state estimate, the objective function is recalculated, and the control sequence is updated continuously using a genetic algorithm. The new control sequence can be optimized by updating the optimal solution of the objective function. During each continuous update, the current system state is used as input to solve the objective function again, generating a new control input sequence.

[0144] This embodiment also provides an intelligent control system for an electrostatic precipitator, including: an electrostatic precipitator modeling module, a deep learning prediction module, and an adaptive optimization module.

[0145] The electrostatic precipitator modeling module is used to construct a mathematical model of the electric field and dust state inside the electrostatic precipitator.

[0146] The deep learning prediction module is used to simplify complex mathematical models and estimate future outputs.

[0147] The adaptive optimization module uses a genetic algorithm to calculate the optimal control sequence and updates the system state based on sensor feedback data.

[0148] This embodiment also provides a computer device applicable to the intelligent control method of an electrostatic precipitator, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the intelligent control method of the electrostatic precipitator as proposed in the above embodiment.

[0149] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0150] This embodiment also provides a storage medium storing a computer program, which, when executed by a processor, implements the intelligent control method for an electrostatic precipitator as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0151] In summary, this invention significantly improves the control accuracy and response stability of electrostatic precipitators by combining deep learning with Kalman filtering. Compared to the frequent adjustments and large fluctuations that occur in traditional control methods, this invention effectively reduces system energy consumption and improves dust removal efficiency by smoothing control increments and optimizing the control sequence. Furthermore, the introduction of a genetic algorithm further enhances the system's adaptability, enabling it to maintain high-efficiency operation under various working conditions.

[0152] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for intelligent control of an electrostatic precipitator, characterized in that: This includes collecting data through sensors, constructing a mathematical model of the electrostatic precipitator process, and simplifying it using a deep learning model to obtain a state-space model; The current state parameters of the state-space model are estimated based on sensor data, and the future output is recursively inferred by combining a deep learning model within the given prediction and control time domains. Define an adaptive weighted quadratic objective function for tracking error and control increment, and output safety constraints; The optimal control parameter sequence is calculated using a genetic algorithm that combines the objective and constraints. First-order control is then executed, and the control parameters are updated on a rolling basis. The method of combining the deep learning model to predict the future output includes fusing the Kalman filter estimate with the prediction of the deep learning model to output the future state parameters. Using data acquired from sensors, the Kalman filter method is used to estimate the current state parameters in the mathematical model, yielding the estimation results. The prior estimates are obtained by the deep learning model based on historical data. Set the current state parameters and control input parameters Input the deep learning model, and through the forward propagation process, output the predicted state parameters for the next time step. The preset prediction and control time domains determine the prediction range of the future output; the future prediction time domain is set to start from the current time. Begin by predicting the future. The state at each time step; during the prediction process, a deep learning model is used to recursively calculate the future. State parameters at time 1 ; The state estimation results of the Kalman filter Predictions from deep learning models Perform weighted fusion to obtain the future state result. ; The adaptive weighted quadratic objective function includes the input state vector of the current system. Define the reference state vector as Indicates the expected target value; Define tracking error , representing state-space parameters With expected target value The difference between them; control increment It is the control input at the current moment. and the control input of the previous moment The amount of change between; The objective function is constructed by including a weighted quadratic term for the tracking error and a weighted quadratic term for the control increment; the weights of the tracking error term are determined by the weighting matrix. The goal of control is to minimize tracking error; The weighted quadratic term of the control increment includes the weights of the control increment term determined by the weighting matrix. Control refers to the degree of penalty imposed on the increment of the control input; the goal is to minimize the change in the control increment. Add control input The constraints make Between the effective range of the control input; for the control increment The constraint is applied, ensuring that the change in the control input is always less than the maximum permissible range. The rolling update of control parameters includes, after obtaining the optimal control sequence, performing first-order control and using the first control input parameter in the control sequence. To control the operation of the electrostatic precipitator, input the data into the electrostatic precipitator's control system, and adjust the system status; After the first-order control is executed, the sensors update the current system state based on the feedback values. After the control operation is executed, the state-space parameters of the electrostatic precipitator change; the system state is then updated based on the data provided by the sensors and the current control strategy. The method of using Kalman filtering algorithm and deep learning model prediction fusion is used to correct the actual observations and predictions. set up This represents the weighting coefficient of the control increment in the objective function, which is adjusted according to the rate of change of the control input; when the change value of the control input is close to the constraint range, the weight is increased to smooth the control process. This represents the tracking error weight in the objective function; if the control error increases with time, then this weight increases. Adjust the value; Based on the latest state estimate Based on the reference trajectory, recalculate the objective function and perform rolling updates of the optimal control sequence. Recalculate the optimal solution of the objective function and update the control strategy; then apply the new system state. As input, the objective function is solved again to generate a new control input sequence.

2. The intelligent control method for an electrostatic precipitator as described in claim 1, characterized in that: The construction of the mathematical model for the electrostatic precipitator process includes constructing a mathematical model of the electric field and dust state inside the electrostatic precipitator. The specific modeling process includes: Assuming the electric field strength of the electrostatic precipitator is a function of space, the electric field distribution is described using the Poisson equation based on the electrode shape and voltage. By applying a voltage between the electrodes of the electrostatic precipitator, the relationship between current density and electric field strength is calculated. The dust state is composed of dust movement and dust concentration; dust particles are affected by gravity, airflow drag force and electric field force, which constitute the motion equation of dust particles; within the working space of the electrostatic precipitator, the change of dust particle concentration is affected by the combined effect of airflow and electric field, and the change of dust concentration is described by the continuous medium equation.

3. The intelligent control method for an electrostatic precipitator as described in claim 2, characterized in that: The simplification using deep learning models includes: pre-training a deep neural network, inputting electric field model parameters and dust state model parameters into the electrostatic precipitator; modeling the relationship between static input and output, and learning the nonlinear mapping between input and output through a multi-layer fully connected network; By extracting nonlinear relationships from the input data, the output obtained by training the deep learning model is used as the state parameters of the state-space model, and the mathematical model is simplified to perform state transitions to obtain the state-space model.

4. The intelligent control method for an electrostatic precipitator as described in claim 3, characterized in that: The process of using a genetic algorithm to calculate the optimal control parameter sequence by combining the objective and constraints includes initializing a genetic algorithm population, where each individual in the population represents a control sequence, and each control input... It is a decision variable within the control time domain; Generate an initial population, where each control input parameter Randomly generated within the control input range; Define a fitness function and use an objective function to evaluate the criteria for each individual's performance; for each individual's sequence... The evaluation is performed using a weighted quadratic objective function with additional constraints. Using the roulette wheel selection method, individuals with high fitness are selected from the current population as parents for crossover and mutation operations; within constraints, the sequence... Fitness weights decrease sequentially. Perform crossover on the selected parent individuals to generate new offspring individuals; select two parent control sequences and perform crossover to generate a new control sequence. Input to the new control sequence Perform random perturbations, and ensure that the mutated control input still meets the control input range constraints and incremental constraints, and iterate to obtain the optimal control input parameter sequence.

5. An intelligent control system for an electrostatic precipitator, based on the intelligent control method for an electrostatic precipitator according to any one of claims 1 to 4, characterized in that: It includes an electrostatic precipitator modeling module, a deep learning prediction module, and an adaptive optimization module; The electrostatic precipitator modeling module is used to construct a mathematical model of the electric field and dust state inside the electrostatic precipitator. The deep learning prediction module is used to simplify complex mathematical models and estimate future outputs. The adaptive optimization module uses a genetic algorithm to calculate the optimal control sequence and updates the system state based on sensor feedback data.

6. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the intelligent control method for the electrostatic precipitator according to any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the intelligent control method for the electrostatic precipitator according to any one of claims 1 to 4.

Images