Cogeneration intelligent optimization method and system based on biomass energy conversion data

Abstract

The application discloses a cogeneration intelligent optimization method and system based on biomass energy conversion data, and relates to the technical field of biomass energy utilization. The method comprises the following steps: collecting and preprocessing key data of a whole biomass energy conversion process and a cogeneration system, and obtaining a standard feature data set; constructing a system mathematical model by fusing a mechanism model and a data-driven model; determining a multi-objective optimization objective function and a constraint condition; performing parameter optimization and solution, and outputting a Pareto optimal parameter solution set; offline verifying the solution set; establishing a dynamic iterative optimization mechanism, updating the model based on real-time data, and triggering dynamic optimization to realize continuous optimization of the system. The system comprises acquisition, main control, optimization and display modules, and corresponding units are arranged under each module, so that the above method can be efficiently realized, and the efficiency of the cogeneration system is effectively improved, the emission of pollutants is reduced, and the stable operation of the system is ensured. The application has the advantages of high efficiency, low emission, high stability and good adaptability.

Description

Technical Field

[0001] This invention relates to the field of biomass energy utilization technology, specifically to a smart optimization method and system for cogeneration based on biomass energy conversion data. Background Technology

[0002] As the global energy structure shifts towards cleaner energy, biomass energy, as a renewable and carbon-neutral energy form, is being increasingly widely used in combined heat and power (CHP). Biomass CHP converts biomass feedstock into electricity and heat, achieving cascaded energy utilization, which is of great significance for alleviating dependence on fossil fuels and reducing carbon emissions.

[0003] Existing intelligent optimization methods for cogeneration often rely on single-mechanism models or data-driven models for system modeling, making it difficult to take into account the strong nonlinearity and complex coupling characteristics of the biomass conversion process. The optimization results deviate significantly from actual operation, and their dynamic adaptability is lacking, failing to respond quickly to real-time operating conditions such as feedstock fluctuations and load changes. This leads to unstable system efficiency and poor pollutant emission control, making it difficult to meet the requirements of high-efficiency, environmentally friendly, and stable operation. Therefore, it is necessary to provide intelligent optimization methods and systems for cogeneration based on biomass energy conversion data to solve the above-mentioned problems. Summary of the Invention

[0004] To address the aforementioned technical problems, this paper provides an intelligent optimization method and system for cogeneration based on biomass energy conversion data. This technical solution solves the problem that existing intelligent optimization methods for cogeneration mentioned in the background technology rely heavily on single mechanism models or data-driven models for system modeling, making it difficult to take into account the strong nonlinearity and complex coupling characteristics of the biomass conversion process. As a result, the optimization results deviate significantly from actual operation, and the methods lack dynamic adaptability, failing to respond quickly to real-time operating conditions such as raw material fluctuations and load changes. This leads to unstable system efficiency, poor pollutant emission control, and difficulty in meeting the requirements for efficient, environmentally friendly, and stable operation.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] Intelligent optimization methods for cogeneration based on biomass energy conversion data include:

[0007] S1. Collect and preprocess key data of the entire biomass energy conversion process and cogeneration system to obtain a standard feature dataset;

[0008] S2. Based on the standard feature dataset, construct a mathematical model for a cogeneration system that integrates the mechanism model and the data-driven model;

[0009] S3. Based on the operational requirements of the cogeneration system, determine the multi-objective optimization objective function and the corresponding constraints;

[0010] S4. Based on the mathematical model of the cogeneration system, the multi-objective optimization objective function and the corresponding constraints, perform parameter optimization and output the Pareto optimal parameter solution set;

[0011] S5. Offline verification of the Pareto optimal parameter solution set is performed by combining simulation verification and experimental verification.

[0012] S6. Establish a dynamic iterative optimization mechanism to update the model based on real-time acquired new data and trigger dynamic optimization to achieve continuous optimization of the cogeneration system.

[0013] In an optional embodiment, the collection and preprocessing of key data from the entire biomass energy conversion process and cogeneration system to obtain a standard feature dataset specifically includes:

[0014] S1.1. Determine the data collection range and simultaneously use a distributed sensor network to collect data on moisture content, cellulose content, hemicellulose content, lignin content, calorific value, and ash content to obtain biomass raw material characteristic data;

[0015] S1.2. Collect pretreatment crushing particle size, pretreatment temperature, pretreatment time, gasification temperature, gasification pressure, equivalence ratio, and residence time to obtain the conversion process parameters;

[0016] S1.3. Collect data on CO content, H2 content, CH4 content, calorific value of fuel gas, ash and slag emissions, and NO content in the gasification gas components. x From the concentrations of SO2 and other concentrations, data on the conversion products were obtained.

[0017] S1.4. Obtain power generation, heat supply, turbine power, heat exchanger efficiency, thermal storage equipment status, and energy storage equipment status through data acquisition terminals to obtain the operating data of the cogeneration system;

[0018] S1.5. Obtain real-time electrical load, real-time heat load, and ambient temperature to determine environmental and demand data;

[0019] S1.6. Perform data cleaning on the biomass raw material characteristic data, conversion process parameters and conversion product data collected above, as well as the cogeneration system operation data, environmental and demand data obtained, including missing value filling and outlier removal;

[0020] S1.7. Standardize the biomass raw material characteristic data, conversion process parameters, conversion product data, cogeneration system operation data, and environmental and demand data after data cleaning, and use the min-max standardization method to convert parameters of different dimensions to the [0, 1] interval;

[0021] S1.8. Perform feature engineering on the standardized biomass feedstock characteristic data, conversion process parameters, conversion product data, cogeneration system operation data, and environmental and demand data to obtain the Pearson correlation coefficient between each characteristic and the overall cogeneration efficiency.

[0022] S1.9. Extract the key features with an absolute value of Pearson correlation coefficient greater than 0.6 from the standardized biomass feedstock characteristic data, conversion process parameters, conversion product data, cogeneration system operation data, and environmental and demand data to obtain the initial feature dataset, including moisture content, gasification temperature, equivalence ratio, gas calorific value, real-time electrical load, and real-time heat load.

[0023] S1.10. Time-align the temporal data in the initial feature dataset to construct a standard feature dataset.

[0024] In an optional embodiment, the step of constructing a mathematical model for a cogeneration system that integrates a mechanistic model and a data-driven model based on a standard feature dataset specifically includes:

[0025] S2.1. Analyze the core unit structure of the combined heat and power system, including the pretreatment unit, conversion unit, power generation unit, heating unit, and thermal / electrical storage unit;

[0026] S2.2. Extract the gasification temperature and equivalence ratio from the standard feature dataset, and obtain the fuel gas composition;

[0027] S2.3. Based on the gasification temperature, equivalence ratio, and fuel composition, a reaction kinetic model is constructed for the conversion unit;

[0028] The reaction kinetic model is expressed as follows:

[0029] In the formula, , and These represent the contents of carbon monoxide, hydrogen, and methane in the fuel gas components, respectively. The gasification temperature of the conversion unit. The equivalent ratio of the conversion units. , and This is a mapping function between gasification temperature, equivalence ratio, and the content of corresponding fuel gas components;

[0030] S2.4. Obtain the condensing temperature and steam temperature of the steam turbine, and establish a steam turbine efficiency formula based on the Rankine cycle for the power generation unit;

[0031] The turbine efficiency formula is as follows:

[0032]

[0033] In the formula, For turbine efficiency, This refers to the condensation temperature of the steam turbine. The steam temperature of the steam turbine;

[0034] S2.5. Select the BP neural network as the basic architecture of the data-driven model to construct the data-driven model;

[0035] S2.6. Establish a fusion mechanism between the mechanistic model and the data-driven model, using the output of the mechanistic model as prior knowledge for the data-driven model, and integrating the outputs through a weighted fusion formula:

[0036]

[0037] In the formula, These are the weighting coefficients. For the output of the mechanistic model, For data-driven model output, Output for the fusion model;

[0038] S2.7. Perform sensitivity analysis on the fusion model to obtain a set of sensitive parameters, and use the set of sensitive parameters as optimization variables for the fusion model to obtain the mathematical model of the cogeneration system.

[0039] In an optional embodiment, determining the multi-objective optimization objective function and corresponding constraints based on the operational requirements of the cogeneration system specifically includes:

[0040] S3.1. Determine the efficiency objective function and define the maximization of overall thermoelectric efficiency:

[0041]

[0042] In the formula, For power generation, To provide heat, The calorific value of biomass feedstock For the quality of biomass raw materials;

[0043] S3.2. Determine the environmental protection objective function and define the minimization of pollutant emissions per unit of thermal power:

[0044]

[0045] In the formula, NO x Emissions SO2 emissions;

[0046] S3.3. Determine the stability objective function and define minimizing the system output fluctuation:

[0047]

[0048] In the formula, The length of the time series. for Output power at any moment Average output power;

[0049] S3.4. Construct a multi-objective optimization function and use a negative sign to transform the maximization problem into a minimization problem:

[0050]

[0051] S3.5. Set equipment constraints, demand constraints, and environmental constraints, and integrate the objective function with the constraints into a mathematical expression.

[0052] In an optional embodiment, the step of performing parameter optimization based on the mathematical model of the cogeneration system, the multi-objective optimization objective function, and the corresponding constraints, and outputting the Pareto optimal parameter solution set, specifically includes:

[0053] S4.1. Select NSGA-II as the intelligent optimization algorithm, and set the population size, maximum number of iterations, crossover probability, and mutation probability;

[0054] S4.2. Define the second optimization variable, including the biomass feed rate. Gasifier equivalence ratio Thermoelectric distribution ratio and the charging and discharging power of thermal storage equipment ;

[0055] S4.3. Initialize the population by randomly generating initial parameter combinations within the feasible region of the variables. Each individual is represented as:

[0056] ;

[0057] S4.4. Input the individual parameters into the mathematical model of the cogeneration system and calculate the values ​​of each objective function:

[0058] ;

[0059] S4.5. Perform non-dominated ranking, divide individuals into different non-dominated layers according to the multi-objective function value, and calculate the crowding distance to assess individual diversity;

[0060] S4.6. Use the tournament selection operator to select parent individuals, and generate child individuals by simulating the binary crossover operator. The formula is:

[0061]

[0062]

[0063] In the formula, , For parent parameters, Cross factor;

[0064] S 4.7. Perform a polynomial mutation operation on the generated offspring individuals, using the following formula:

[0065]

[0066] In the formula, As a variable factor, , These are the maximum and minimum values ​​of the variable, respectively.

[0067] S4.8. Merge the parent and offspring populations, re-perform the non-dominated sorting and crowding calculation, retain the best individual for the next iteration, until the maximum number of iterations is reached, and output the final Pareto optimal parameter solution set.

[0068] In an optional embodiment, the establishment of a dynamic iterative optimization mechanism, which updates the model based on real-time acquired new data and triggers dynamic optimization to achieve continuous optimized operation of the cogeneration system, specifically includes:

[0069] S6.1. Deploy real-time data acquisition terminals to collect real-time key data of biomass raw materials once per hour and store them in a real-time database;

[0070] S6.2. Set the model update trigger conditions to trigger the model update process;

[0071] S6.3. Use the incremental SVM algorithm to update the data-driven model, training only with new data, and updating the weights using the following formula:

[0072]

[0073] In the formula, The original weights, For learning rate, For the actual output, To predict the output, Input features;

[0074] S6.4. Execute dynamic optimization trigger judgment, check whether the current parameters deviate from the optimal solution, if so, start the NSGA-II algorithm for incremental optimization.

[0075] S6.5. Dynamically adjust and optimize target weights, increasing the weight of efficiency targets during peak electricity consumption periods and increasing the weight of environmental protection targets during environmentally sensitive periods. The weight adjustment formula is as follows:

[0076]

[0077]

[0078] In the formula Weights for efficiency objectives Weighting for environmental protection goals, For real-time electrical load, This is the maximum electrical load;

[0079] S6.6. Send the new optimized parameters to the control system and achieve real-time adjustment through the PID controller;

[0080] S6.7. Record the changes in performance indicators before and after optimization, including efficiency improvement, reduction in pollutant emissions, and reduction in output fluctuation, and generate an optimization effect report.

[0081] S6.8. Based on the optimization effect report, evaluate the dynamic iteration mechanism and calculate the optimization stability index:

[0082]

[0083] In the formula, For the standard deviation of efficiency fluctuation, This represents the average efficiency.

[0084] Furthermore, a smart optimization system for cogeneration based on biomass energy conversion data is proposed to implement the optimization method described above, characterized by comprising:

[0085] The data acquisition module is used to collect key data from the entire biomass energy conversion process and also to collect key data from the cogeneration system.

[0086] The main control module is used to receive data transmitted by the acquisition module, to preprocess key data of the entire biomass energy conversion process and cogeneration system, to obtain standard feature datasets, to construct a mathematical model of cogeneration system that integrates mechanism model and data-driven model based on the standard feature datasets, and to determine multi-objective optimization objective function and corresponding constraints according to the operation requirements of cogeneration system.

[0087] The optimization module is used to solve for parameters based on the mathematical model of the cogeneration system, the multi-objective optimization objective function and the corresponding constraints, and output the Pareto optimal parameter solution set. It is used to verify the Pareto optimal parameter solution set offline by combining simulation verification and experimental verification, and to establish a dynamic iterative optimization mechanism to update the model based on the new data collected in real time and trigger dynamic optimization to realize the continuous optimized operation of the cogeneration system.

[0088] The display module is used to present the process and results of intelligent optimization of cogeneration to the user.

[0089] In an optional embodiment, the acquisition module includes:

[0090] The first acquisition unit is used to collect key data from the entire biomass energy conversion process.

[0091] The second acquisition unit is used to acquire key data from the combined heat and power system.

[0092] In an optional embodiment, the main control module includes:

[0093] The data receiving unit is used to receive data transmitted by the acquisition module;

[0094] The data processing unit is used to preprocess key data of the entire biomass energy conversion process and cogeneration system to obtain a standard feature dataset.

[0095] The model building unit is used to build a mathematical model of a cogeneration system that integrates a mechanism model and a data-driven model based on a standard feature dataset.

[0096] The model constraint unit is used to determine the multi-objective optimization objective function and corresponding constraint conditions based on the operational requirements of the cogeneration system.

[0097] In an optional embodiment, the optimization module includes:

[0098] The optimization solution unit is used to perform parameter optimization based on the mathematical model of the cogeneration system, the multi-objective optimization objective function and the corresponding constraints, and output the Pareto optimal parameter solution set.

[0099] An offline verification unit is used to perform offline verification of the Pareto optimal parameter solution set by combining simulation verification and experimental verification.

[0100] The dynamic optimization unit is used to establish a dynamic iterative optimization mechanism, update the model based on real-time acquired new data and trigger dynamic optimization to achieve continuous optimized operation of the cogeneration system.

[0101] Compared with the prior art, the beneficial effects of the present invention are:

[0102] The proposed intelligent optimization method for cogeneration based on biomass energy conversion data collects and preprocesses key data from the entire biomass energy conversion process and the cogeneration system, constructs a standard feature dataset, and integrates mechanistic models and data-driven models to establish a system mathematical model. This method accurately characterizes the strong nonlinearity and complex coupling characteristics of the biomass conversion process, provides a reliable quantitative tool for optimization decisions, and effectively improves the model's adaptability to actual operating conditions.

[0103] The proposed intelligent optimization method for cogeneration based on biomass energy conversion data determines the multi-objective optimization objective function and constraints, uses the NSGA-II algorithm to optimize parameters and output the Pareto optimal parameter solution set, and achieves coordinated optimization of cogeneration efficiency, pollutant emissions and system stability. Under the premise of ensuring equipment safety and meeting load requirements, it can find the optimal operating scheme that balances efficiency, environmental protection and stability, and improve the overall operating performance of the system.

[0104] The proposed intelligent optimization method for cogeneration based on biomass energy conversion data establishes a dynamic iterative optimization mechanism. By updating the model based on real-time data and triggering dynamic optimization, the cogeneration system can quickly respond to real-time operating conditions such as fluctuations in feedstock characteristics and load changes. It can continuously output optimized parameters that are adapted to the current state, ensuring that the system maintains efficient, environmentally friendly, and stable operation in the long term, and enhancing the system's adaptability and robustness under complex operating conditions. Attached Figure Description

[0105] Figure 1 This is a flowchart of the intelligent optimization method for cogeneration based on biomass energy conversion data proposed in this invention;

[0106] Figure 2 This is a flowchart illustrating the construction of the mathematical model for the cogeneration system in this invention.

[0107] Figure 3 This is a flowchart illustrating the process of obtaining the Pareto optimal parameter solution set in this invention.

[0108] Figure 4 This is a system framework diagram of the intelligent optimization system for cogeneration based on biomass energy conversion data proposed in this invention. Detailed Implementation

[0109] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0110] Reference Figure 1 - Figure 4 As shown, the intelligent optimization method for cogeneration based on biomass energy conversion data includes:

[0111] S1. Collect and preprocess key data of the entire biomass energy conversion process and cogeneration system to obtain a standard feature dataset;

[0112] S2. Based on the standard feature dataset, construct a mathematical model for a cogeneration system that integrates the mechanism model and the data-driven model;

[0113] S3. Based on the operational requirements of the cogeneration system, determine the multi-objective optimization objective function and the corresponding constraints;

[0114] S4. Based on the mathematical model of the cogeneration system, the multi-objective optimization objective function and the corresponding constraints, perform parameter optimization and output the Pareto optimal parameter solution set;

[0115] S5. Offline verification of the Pareto optimal parameter solution set is performed by combining simulation verification and experimental verification.

[0116] S6. Establish a dynamic iterative optimization mechanism to update the model based on real-time acquired new data and trigger dynamic optimization to achieve continuous optimization of the cogeneration system.

[0117] Furthermore, key data from the entire biomass energy conversion process and cogeneration system are collected and preprocessed to obtain a standard feature dataset, specifically including:

[0118] S1.1. Determine the data collection range and simultaneously use a distributed sensor network to collect data on moisture content, cellulose content, hemicellulose content, lignin content, calorific value, and ash content to obtain biomass raw material characteristic data;

[0119] S1.2. Collect pretreatment crushing particle size, pretreatment temperature, pretreatment time, gasification temperature, gasification pressure, equivalence ratio, and residence time to obtain the conversion process parameters;

[0120] S1.3. Collect data on CO content, H2 content, CH4 content, calorific value of fuel gas, ash and slag emissions, and NO content in the gasification gas components. x From the concentrations of SO2 and other concentrations, data on the conversion products were obtained.

[0121] S1.4. Obtain power generation, heat supply, turbine power, heat exchanger efficiency, thermal storage equipment status, and energy storage equipment status through data acquisition terminals to obtain the operating data of the cogeneration system;

[0122] S1.5. Obtain real-time electrical load, real-time heat load, and ambient temperature to determine environmental and demand data;

[0123] S1.6. Perform data cleaning on the biomass raw material characteristic data, conversion process parameters and conversion product data collected above, as well as the cogeneration system operation data, environmental and demand data obtained, including missing value filling and outlier removal;

[0124] S1.7. Standardize the biomass raw material characteristic data, conversion process parameters, conversion product data, cogeneration system operation data, and environmental and demand data after data cleaning, and use the min-max standardization method to convert parameters of different dimensions to the [0, 1] interval;

[0125] S1.8. Perform feature engineering on the standardized biomass feedstock characteristic data, conversion process parameters, conversion product data, cogeneration system operation data, and environmental and demand data to obtain the Pearson correlation coefficient between each characteristic and the overall cogeneration efficiency.

[0126] S1.9. Extract the key features with an absolute value of Pearson correlation coefficient greater than 0.6 from the standardized biomass feedstock characteristic data, conversion process parameters, conversion product data, cogeneration system operation data, and environmental and demand data to obtain the initial feature dataset, including moisture content, gasification temperature, equivalence ratio, gas calorific value, real-time electrical load, and real-time heat load.

[0127] S1.10. Time-align the temporal data in the initial feature dataset to construct a standard feature dataset.

[0128] Specifically, the data collection scope covers biomass feedstock characteristic data, conversion process parameters, conversion product data, cogeneration system operation data, and environmental and demand data. When collecting this data using a distributed sensor network and data acquisition terminals, the collection frequency is set to once per hour, with a high-frequency collection mode of 10 minutes per collection for key parameters such as gasification temperature and equivalence ratio. The raw data is cleaned by using linear interpolation to fill in missing values. For example, if a certain type of data in the original data... middle If missing, then Outlier removal is based on the 3σ criterion. Data values ​​exceeding the range of [μ-3σ, μ+3σ] are identified as outliers and removed, where μ is the data mean and σ is the standard deviation. Key features with an absolute Pearson correlation coefficient greater than 0.6 are selected, including moisture content, gasification temperature, equivalence ratio, fuel gas calorific value, real-time electrical load, and real-time heat load. Time-alignment is performed on the temporal data in the initial feature dataset to construct a standard feature dataset. This time-alignment, using timestamps as a benchmark, matches the conversion process parameters with the corresponding thermoelectric output data in the time dimension, ensuring accurate correlation of thermoelectric output data within one hour after a change in gasification parameters.

[0129] Understandably, before constructing a standard feature dataset, it is necessary to conduct a quality assessment of the feature dataset. The data integrity index I_com = (total data volume - missing data volume) / total data volume is calculated, and I_com ≥ 95% is required. The data accuracy index I_acc = 1 - |measured value - standard value| / standard value is calculated, and I_acc ≥ 90% is required. If the index requirements are met, a standardized feature dataset is output.

[0130] Furthermore, based on a standard feature dataset, a mathematical model for a combined heat and power system is constructed that integrates a mechanistic model and a data-driven model, specifically including:

[0131] S2.1. Analyze the core unit structure of the combined heat and power system, including the pretreatment unit, conversion unit, power generation unit, heating unit, and thermal / electrical storage unit;

[0132] S2.2. Extract the gasification temperature and equivalence ratio from the standard feature dataset, and obtain the fuel gas composition;

[0133] S2.3. Based on the gasification temperature, equivalence ratio, and fuel composition, a reaction kinetic model is constructed for the conversion unit;

[0134] The reaction kinetics model is expressed as follows:

[0135]

[0136] In the formula, , and These represent the contents of carbon monoxide, hydrogen, and methane in the fuel gas components, respectively. The gasification temperature of the conversion unit. The equivalent ratio of the conversion units. , and This is a mapping function between gasification temperature, equivalence ratio, and the content of corresponding fuel gas components;

[0137] S2.4. Obtain the condensing temperature and steam temperature of the steam turbine, and establish a steam turbine efficiency formula based on the Rankine cycle for the power generation unit;

[0138] The formula for turbine efficiency is:

[0139]

[0140] In the formula, For turbine efficiency, This refers to the condensation temperature of the steam turbine. The steam temperature of the steam turbine;

[0141] S2.5. Select the BP neural network as the basic architecture of the data-driven model to construct the data-driven model;

[0142] S2.6. Establish a fusion mechanism between the mechanistic model and the data-driven model, using the output of the mechanistic model as prior knowledge for the data-driven model, and integrating the outputs through a weighted fusion formula:

[0143]

[0144] In the formula, The weighting coefficients and , For the output of the mechanistic model, For data-driven model output, Output for the fusion model;

[0145] S2.7. Perform sensitivity analysis on the fusion model to obtain the set of sensitive parameters. The set of sensitive parameters is used as the optimization variables of the fusion model to obtain the mathematical model M(X,Y) of the cogeneration system, where X is the input parameter vector and Y is the output performance vector.

[0146] Specifically, analyzing the core unit structure of a combined heat and power (CHP) system involves clarifying the process relationships from biomass feedstock to pretreatment unit, conversion unit, power generation unit, heating unit, and thermal / electrical storage unit, and defining the input, output, and intermediate variables for each unit. This is reflected in the expression formulas of the reaction kinetic model. , and : Represents the content of carbon monoxide (CO), hydrogen (H2), and methane (CH4) in the gasified gas (commonly expressed as volume fraction, mole fraction, or mass fraction, depending on the process and testing standards), which are key indicators for measuring the quality of gasification products and determining the efficiency of subsequent cogeneration. T2: Refers to the gasification temperature of the gasification unit (when the conversion unit uses gasification technology), which is a core thermal environment parameter for the thermochemical reaction of biomass, affecting the reaction rate and the composition of the products. λ: Equivalent ratio, reflecting the ratio of oxidant (such as air, oxygen, etc.) to biomass feedstock during gasification, determining the oxidation-reduction atmosphere of the reaction, and playing a key regulatory role in the composition of gasification products. The mapping relationship function is an empirical formula derived from the mechanism. Since biomass gasification involves multiple steps such as cracking, oxidation, and reduction, taking the main reaction of carbon with oxygen and water vapor as an example (simplified), the relationship between components and parameters can be derived:

[0147] Carbon monoxide formation: If we focus on the dominant stage of incomplete oxidation and reduction reactions of carbon, assuming the main reaction is...

[0148] C + H₂O = CO + H₂, C + CO₂ = 2CO. Combining the reaction equilibrium constant (affected by temperature T₂) and the conservation of mass (related to the equivalence ratio λ), an empirical function similar to the Arrhenius form can be constructed:

[0149] ;

[0150] In the formula, Pre-exponential factors (related to raw material characteristics, reactor structure, etc.) The activation energy of the reaction. The gas constant is... Let be the reaction order. It is the equivalence ratio threshold where carbon monoxide formation is dominant. Below this value, the oxidation reaction is limited, while above this value, excessive oxygen may lead to deep oxidation.

[0151] Hydrogen generation: If the reduction reactions involving water vapor, C + H₂O = CO + H₂ and CO + H₂O = CO₂ + H₂, are the main components, considering the effect of temperature on the reaction equilibrium (high temperature favors hydrogen generation) and the influence of the stoichiometric ratio (lower λ results in a stronger reducing atmosphere), the functional form is:

[0152] ;

[0153] in, , , , To fit the parameters to the raw materials and processes, reflecting the complex effects of temperature promotion and equivalence ratio suppression (because a high λ will introduce more oxygen and consume hydrogen);

[0154] Methane formation: Methane mainly originates from biomass pyrolysis and subsequent hydrogenation reactions. Low temperatures and reducing atmospheres favor its formation (high temperatures easily decompose it into CO and H2). The function may show a negative correlation with temperature and a positive correlation with equivalence ratio (lower λ corresponds to a reducing atmosphere).

[0155]

[0156] when Excess oxygen disrupts the reducing environment, thus inhibiting methane formation. Approaching 0, , , , It needs to be calibrated through experiments to reflect characteristics such as the volatile content and hydrogenation reactivity of the raw materials.

[0157] Specifically, a backpropagation (BP) neural network was selected as the basic architecture of the data-driven model. The input layer included key features such as biomass moisture content ω, gasification temperature T2, and equivalence ratio λ. The hidden layer consisted of three layers, with each layer having 1.5 times the number of neurons as the input features. The output layer represented the overall thermoelectric efficiency η_total. Training and testing sets were then created, with samples drawn from the standardized feature dataset at a 7:3 ratio. The BP neural network was trained using the training set, and the weight parameters were adjusted using the backpropagation algorithm. The objective function was to minimize the prediction error.

[0158]

[0159] In the formula, This represents the prediction error (or mean squared error), used to quantify the overall deviation between the model's predictions and the actual values. The smaller the value, the higher the model's prediction accuracy. The sample size refers to the total number of data points in the dataset used to calculate the error, reflecting the statistical sample size. The model prediction value for the i-th sample is the prediction result (such as the combined heat and power efficiency, power generation, etc.) output after inputting biomass parameters, operating conditions, etc., through the mathematical model of the combined heat and power system (such as the model that integrates mechanism and data-driven model). The actual observed value of the i-th sample is the real data (corresponding to the same index) collected in an actual cogeneration system through sensors, experimental detection, and other means. Then, a test set is used to verify the model accuracy, and the root mean square error is calculated.

[0160]

[0161] The requirement is RMSE ≤ 5%. If this requirement is not met, the number of training iterations should be increased or the network structure should be adjusted.

[0162] Furthermore, based on the operational requirements of the combined heat and power system, the multi-objective optimization objective function and corresponding constraints are determined, specifically including:

[0163] S3.1. Determine the efficiency objective function and define the maximization of overall thermoelectric efficiency:

[0164]

[0165] In the formula, For power generation, To provide heat, The calorific value of biomass feedstock For the quality of biomass raw materials;

[0166] S3.2. Determine the environmental protection objective function and define the minimization of pollutant emissions per unit of thermal power:

[0167]

[0168] In the formula, NO x Emissions SO2 emissions;

[0169] S3.3. Determine the stability objective function and define minimizing the system output fluctuation:

[0170]

[0171] In the formula, The length of the time series. for Output power at any moment Average output power;

[0172] S3.4. Construct a multi-objective optimization function and use a negative sign to transform the maximization problem into a minimization problem:

[0173]

[0174] S3.5. Set equipment constraints, demand constraints, and environmental constraints, and integrate the objective function with the constraints into a mathematical expression.

[0175] Specifically, the equipment constraints can be set by professionals based on experience, such as gasifier temperature T2 ≤ 1200℃, turbine power P_t ≤ 1.1 × P_rated, where P_rated is the rated power; raw material constraints m ≤ m_max, where m_max is the maximum daily raw material supply. Demand constraints are set: heat supply H ≥ H_min, power generation E ≥ E_min, where H_min is the minimum heat load and E_min is the minimum electrical load. Environmental constraints are set, such as NO... x Emissions ≤ 50 mg / m³, SO₂ emissions ≤ 35 mg / m³, meeting the requirements of national standard GB 13223-2011. The objective function and constraints are integrated into a mathematical expression: min F, st T² ≤ 1200℃, P_t ≤ 1.1 × P_rated, m ≤ m_max, H ≥ H_min, E ≥ E_min, NO x ≤50mg / m³, SO2≤35mg / m³.

[0176] Furthermore, based on the mathematical model of the cogeneration system, the multi-objective optimization objective function, and the corresponding constraints, parameter optimization is performed to output the Pareto optimal parameter solution set, specifically including:

[0177] S4.1. Select NSGA-II as the intelligent optimization algorithm, and set the population size, maximum number of iterations, crossover probability, and mutation probability;

[0178] S4.2. Define the second optimization variable, including the biomass feed rate. Gasifier equivalence ratio Thermoelectric distribution ratio ( ) and the charging and discharging power of thermal storage equipment ;

[0179] S4.3. Initialize the population by randomly generating initial parameter combinations within the feasible region of the variables. Each individual is represented as:

[0180] ;

[0181] S4.4. Input the individual parameters into the mathematical model of the cogeneration system and calculate the values ​​of each objective function:

[0182] ;

[0183] S4.5. Perform non-dominated ranking, divide individuals into different non-dominated layers according to the multi-objective function value, and calculate the crowding distance to assess individual diversity;

[0184] S4.6. Use the tournament selection operator to select parent individuals, and generate child individuals by simulating the binary crossover operator. The formula is:

[0185]

[0186]

[0187] In the formula, , The parameter values ​​representing offspring individuals are new candidate solutions generated after crossover operations, corresponding to operating parameters (such as gasification temperature, feed rate, etc.) in cogeneration optimization. , For parent parameters, As a crossover factor, it controls the "divergence" of the crossover operation and is a key factor in adjusting the differences in parameters between offspring and parents;

[0188] S 4.7. Perform a polynomial mutation operation on the generated offspring individuals, using the following formula:

[0189]

[0190] In the formula, As a variable factor, , These are the maximum and minimum values ​​of the variable, respectively.

[0191] S4.8. Combine the parent and offspring populations, re - perform non - dominated sorting and crowding degree calculation, retain the optimal individuals to enter the next iteration until the maximum number of iterations is reached, and output the final Pareto - optimal parameter solution set.

[0192] Specifically, set the population size to 100, the maximum number of iterations to 1000, the crossover probability to 0.9, and the mutation probability to 0.05. Population size 100: Balance the solution space coverage and computational cost, and adapt to the search requirements of 4 - 6 - dimensional variables; Maximum number of iterations 1000: Ensure that the algorithm converges to a stable state and covers the key areas of the Pareto front; Crossover probability 0.9: Promote the fusion of high - quality genes and maintain the population diversity of multi - objective optimization; Mutation probability 0.05: Introduce appropriate perturbations to avoid premature convergence and improve the global optimality of the solutions. In step S4.8, combine the parent and offspring populations, re - perform non - dominated sorting and crowding degree calculation, and retain the top 100 optimal individuals to enter the next iteration until the maximum number of iterations is reached, and output the final Pareto - optimal parameter solution set. In step S4.3, the variable feasible region refers to the value range of the optimization variables (biomass feeding rate m in , gasifier equivalence ratio λ, thermoelectric distribution ratio k, charge - discharge power P of the heat storage device heat ). This range is jointly determined by the system physical constraints, equipment performance limitations, and actual operation requirements (set by the operator) to ensure the feasibility of the parameter combination in engineering practice. Specifically, the feasible regions of each variable should meet: The biomass feeding rate should not exceed the device design processing capacity (such as the daily maximum raw material supply) and should not be lower than the minimum feeding rate to maintain the stable operation of the system; The gasifier equivalence ratio should be in the interval to ensure conversion efficiency and product quality (such as the gasification process is usually 0.2 - 0.4); The thermoelectric distribution ratio should meet the demand balance of real - time electrical load and heat load (such as 0 < k < 1); The charge - discharge power of the heat storage device should not exceed the device rated power and the heat storage capacity limit. The setting of these ranges aims to prevent parameters from exceeding the equipment safe operation threshold or causing system performance deterioration, and provide reasonable parameter boundaries for population initialization.

[0193] Understandably, in step S4.5, non-dominated sorting is performed, and individuals are divided into different non-dominated layers based on the multi-objective function values. Crowding distance is calculated to assess individual diversity. That is, based on each objective function value, individuals not dominated by other individuals are grouped into the same non-dominated layer, and the uniformity of population distribution is assessed by calculating the distance between each individual and its neighbors in the objective space. In step S4.6, a tournament selection operator is used to select parent individuals, and a simulated binary crossover operator is used to generate offspring individuals. First, a certain number of individuals are randomly selected from the population, and the better ones are chosen as parents; then, offspring are generated using a formula. In step S4.8, the parent and offspring populations are merged, and non-dominated sorting and crowding calculation are performed again. The best individuals are retained for the next iteration until the maximum number of iterations is reached, and the final Pareto optimal parameter solution set is output. Non-dominated ranking is a rule used in multi-objective optimization (such as cogeneration, which requires simultaneous optimization of efficiency, environmental protection, and cost) to distinguish the "superiority levels" of candidate solutions. Non-dominated ranking involves finding all "top-level solutions" (first level) that are not dominated by any other solution, then finding new top-level solutions (second level) from the remaining solutions, and so on, thus classifying all candidate solutions into levels. For example, suppose the optimization objective is "maximum efficiency, lowest emissions," and there are 3 candidate solutions:

[0194] Solution 1: Efficiency 80%, emissions 100mg / m³

[0195] Solution 2: Efficiency 85%, emissions 95mg / m³

[0196] Solution 3: Efficiency 75%, emissions 90mg / m³

[0197] Sorting process:

[0198] Solution 2 has higher efficiency than both solutions 1 and 3, and lower emissions than solution 1. Therefore, **solution 2 dominates solution 1**. However, solution 2's emissions are greater than solution 3's, and solution 3's efficiency is less than solution 2's. Thus, solutions 2 and 3 do not dominate each other.

[0199] Solution 3 has the lowest emissions, but its efficiency is not the highest → it does not dominate any solution;

[0200] Ultimately: the first level (highest quality) consists of solutions 2 and 3 (which do not dominate each other and are not dominated by other solutions); the second level consists of solution 1 (which is dominated by solution 2).

[0201] Crowding is an indicator that measures the "crowding" around a candidate solution, reflecting the "uniformity of distribution" of solutions in the target space. Simply put:

[0202] For each objective (such as efficiency or emissions), calculate the distance between the solution and its neighboring solutions;

[0203] Add up the distances of all targets to get the "crowding" of the solution - the smaller the value, the denser the surrounding solutions; the larger the value, the "lonely high-quality solution" (with few competitors around).

[0204] Suppose that there is a solution i in the objective space, and its neighboring solutions in the efficiency objective are i. left (Less efficient) and i right (More efficient), the adjacent solution to the emission target is j left (Higher emissions) and j right (Lower emissions) then congestion level I crowd Calculate: I crowd = (Efficiency target distance / Efficiency target range) + (Emission target distance / Emission target range).

[0205] Efficiency target distance = i right Efficiency-i lef efficiency;

[0206] Emission target distance = j right Emissions - j left emission;

[0207] Target range = maximum value of the target - minimum value (e.g., efficiency range 80%-90%, emission range 80-120 mg / m³).

[0208] Furthermore, offline verification of the Pareto optimal parameter solution set is conducted through a combination of simulation and experimental verification, specifically including:

[0209] S5.1. Build a simulation verification platform, input each set of parameters in the Pareto optimal parameter solution set into the mathematical model of the cogeneration system, simulate the system operation state under different parameter combinations, output key performance indicators such as power generation, heat supply, pollutant emissions, and efficiency, record the simulation results and analyze the mapping relationship between parameters and performance, and select the parameter subset with the best performance at the simulation level.

[0210] S5.2. Select parameter combinations that perform well in simulation verification and conduct experimental verification in an actual cogeneration test system. Set operating parameters such as biomass feed rate, gasification temperature, and equivalence ratio according to the selected parameters, collect real-time operating data of the system, including the gas composition of the conversion unit, the turbine power of the power generation unit, the heat output of the heating unit, and the pollutant emission concentration, and compare them with the simulation results.

[0211] S5.3. Calculate the deviation between the simulation value and the experimental value, and use indicators such as root mean square error and mean absolute error to evaluate the prediction accuracy of the model. If the deviation is within the allowable range (e.g., ≤5%), it indicates that the parameter solution set is reliable. If the deviation is too large, backtrack to the parameter optimization stage, check whether the model assumptions or constraints are reasonable, correct them, regenerate the Pareto optimal parameter solution set, and verify it again.

[0212] S5.4. Based on the comprehensive simulation and experimental verification results, parameter combinations that do not meet the actual operating requirements are eliminated, and parameter solution sets that combine theoretical optimization and engineering feasibility are retained, so as to provide a reliable initial benchmark for subsequent dynamic iterative optimization.

[0213] Furthermore, a dynamic iterative optimization mechanism is established to update the model based on real-time acquired new data and trigger dynamic optimization, thereby achieving continuous optimization of the cogeneration system. Specifically, this includes:

[0214] S6.1. Deploy real-time data acquisition terminals to collect real-time key data of biomass raw materials once per hour and store them in a real-time database;

[0215] S6.2. Set the model update trigger conditions to trigger the model update process;

[0216] S6.3. Use the incremental SVM algorithm to update the data-driven model, training only with new data, and updating the weights using the following formula:

[0217]

[0218] In the formula, The original weights, For learning rate, For the actual output, To predict the output, Input features;

[0219] S6.4. Execute dynamic optimization trigger judgment, check whether the current parameters deviate from the optimal solution, if so, start the NSGA-II algorithm for incremental optimization.

[0220] S6.5. Dynamically adjust and optimize target weights, increasing the weight of efficiency targets during peak electricity consumption periods and increasing the weight of environmental protection targets during environmentally sensitive periods. The weight adjustment formula is as follows:

[0221]

[0222]

[0223] In the formula Weights for efficiency objectives Weighting for environmental protection goals, For real-time electrical load, This is the maximum electrical load;

[0224] S6.6. Send the new optimized parameters to the control system and achieve real-time adjustment through the PID controller;

[0225] S6.7. Record the changes in performance indicators before and after optimization, including efficiency improvement, reduction in pollutant emissions, and reduction in output fluctuation, and generate an optimization effect report.

[0226] S6.8. Based on the optimization effect report, evaluate the dynamic iteration mechanism and calculate the optimization stability index:

[0227]

[0228] In the formula, For the standard deviation of efficiency fluctuation, This represents the average efficiency.

[0229] Specifically, the model update trigger condition is set so that the model update process is triggered when the change in raw material moisture content Δω is ≥10% or the load fluctuation ΔL is ≥15% for three consecutive hours. In step S6.4, a dynamic optimization trigger judgment is performed every 2 hours to check whether the current parameters deviate from the optimal solution by more than 5%. If so, the NSGA-II algorithm is started for incremental optimization. In step S6.6, the new optimized parameters are sent to the control system, and real-time adjustment is achieved through the PID controller, with a control deviation ≤2%.

[0230] Furthermore, a smart optimization system for cogeneration based on biomass energy conversion data is proposed to implement the optimization method described above, characterized by comprising:

[0231] The data acquisition module is used to collect key data from the entire biomass energy conversion process, as well as key data from the cogeneration system.

[0232] The main control module receives data transmitted from the acquisition module, preprocesses key data of the entire biomass energy conversion process and cogeneration system, obtains standard feature datasets, constructs a mathematical model of cogeneration system based on the standard feature datasets that integrates the mechanism model and the data-driven model, and determines the multi-objective optimization objective function and corresponding constraints according to the operating requirements of the cogeneration system.

[0233] The optimization module is used to solve for parameters based on the mathematical model of the cogeneration system, the multi-objective optimization objective function and the corresponding constraints, and output the Pareto optimal parameter solution set. It is used to verify the Pareto optimal parameter solution set offline by combining simulation verification and experimental verification, and to establish a dynamic iterative optimization mechanism. It updates the model based on the new data collected in real time and triggers dynamic optimization to achieve continuous optimization of the cogeneration system.

[0234] The display module is used to present the process and results of intelligent optimization of cogeneration to the user.

[0235] Furthermore, the data acquisition module includes:

[0236] The first data acquisition unit is used to collect key data from the entire biomass energy conversion process.

[0237] The second acquisition unit is used to acquire key data from the combined heat and power system.

[0238] Furthermore, the main control module includes:

[0239] The data receiving unit is used to receive data transmitted by the acquisition module.

[0240] The data processing unit is used to preprocess key data of the entire biomass energy conversion process and cogeneration system to obtain standard feature datasets.

[0241] The model building unit is used to construct a mathematical model of a cogeneration system that integrates a mechanistic model and a data-driven model based on a standard feature dataset.

[0242] The model constraint unit is used to determine the multi-objective optimization objective function and corresponding constraint conditions based on the operational requirements of the cogeneration system.

[0243] Furthermore, the optimization modules include:

[0244] The optimization solution unit is used to perform parameter optimization based on the mathematical model of the cogeneration system, the multi-objective optimization objective function and the corresponding constraints, and output the Pareto optimal parameter solution set.

[0245] The offline verification unit is used to perform offline verification of the Pareto optimal parameter solution set by combining simulation verification and experimental verification.

[0246] The dynamic optimization unit is used to establish a dynamic iterative optimization mechanism, which updates the model based on new data collected in real time and triggers dynamic optimization to achieve continuous optimization of the cogeneration system.

[0247] The advantages of this invention are as follows: Through full-process data acquisition, preprocessing, and fusion modeling, it accurately depicts the complex characteristics of biomass conversion, improving model adaptability; it employs a multi-objective optimization algorithm to achieve synergistic optimization of efficiency, environmental protection, and stability, ensuring efficient system operation under safety constraints; relying on a dynamic iteration mechanism, it responds in real-time to raw material fluctuations and load changes, continuously outputting optimal parameters to ensure long-term stability and high efficiency. The modular design of the system ensures close integration of each link, balancing data reliability, model accuracy, and operating condition adaptability, comprehensively improving the energy utilization efficiency and environmental performance of biomass cogeneration.

[0248] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A smart optimization method for cogeneration based on biomass energy conversion data, characterized in that, include: S1. Collect and preprocess key data of the entire biomass energy conversion process and cogeneration system to obtain a standard feature dataset; S2. Based on the standard feature dataset, construct a mathematical model for a cogeneration system that integrates the mechanism model and the data-driven model; S3. Based on the operational requirements of the cogeneration system, determine the multi-objective optimization objective function and the corresponding constraints; S4. Based on the mathematical model of the cogeneration system, the multi-objective optimization objective function and the corresponding constraints, perform parameter optimization and output the Pareto optimal parameter solution set; S5. Offline verification of the Pareto optimal parameter solution set is performed by combining simulation verification and experimental verification. S6. Establish a dynamic iterative optimization mechanism to update the model based on real-time new data and trigger dynamic optimization to achieve continuous optimized operation of the cogeneration system; The mathematical model for the cogeneration system, which integrates the mechanistic model and the data-driven model based on the standard feature dataset, specifically includes: S2.

1. Analyze the core unit structure of the combined heat and power system, including the pretreatment unit, conversion unit, power generation unit, heating unit, and thermal / electrical storage unit; S2.

2. Extract the gasification temperature and equivalence ratio from the standard feature dataset, and obtain the fuel gas composition; S2.

3. Based on the gasification temperature, equivalence ratio, and fuel composition, a reaction kinetic model is constructed for the conversion unit; The reaction kinetic model is expressed as follows: In the formula, , and These represent the contents of carbon monoxide, hydrogen, and methane in the fuel gas components, respectively. The gasification temperature of the conversion unit. The equivalent ratio of the conversion units. , and This is a mapping function between gasification temperature, equivalence ratio, and the content of corresponding fuel gas components; S2.

4. Obtain the condensing temperature and steam temperature of the steam turbine, and establish a steam turbine efficiency formula based on the Rankine cycle for the power generation unit; The turbine efficiency formula is as follows: In the formula, For turbine efficiency, This refers to the condensation temperature of the steam turbine. The steam temperature of the steam turbine; S2.

5. Select the BP neural network as the basic architecture of the data-driven model to construct the data-driven model; S2.

6. Establish a fusion mechanism between the mechanistic model and the data-driven model, using the output of the mechanistic model as prior knowledge for the data-driven model, and integrating the outputs through a weighted fusion formula: In the formula, These are the weighting coefficients. For the output of the mechanistic model, For data-driven model output, Output for the fusion model; S2.

7. Perform sensitivity analysis on the fusion model to obtain the set of sensitive parameters. ,in, The biomass moisture content is used as the optimization variable for the fusion model, thereby obtaining the mathematical model of the cogeneration system; The process of determining the multi-objective optimization objective function and corresponding constraints based on the operational requirements of the combined heat and power system specifically includes: S3.

1. Determine the efficiency objective function and define the maximization of overall thermoelectric efficiency: In the formula, For power generation, To provide heat, The calorific value of biomass feedstock For the quality of biomass raw materials; S3.

2. Determine the environmental protection objective function and define the minimization of pollutant emissions per unit of thermal power: In the formula, NO x Emissions SO2 emissions; S3.

3. Determine the stability objective function and define minimizing the system output fluctuation: In the formula, The length of the time series. for Output power at any moment Average output power; S3.

4. Construct a multi-objective optimization function and use a negative sign to transform the maximization problem into a minimization problem: S3.

5. Set equipment constraints, demand constraints, and environmental constraints, and integrate the objective function and constraints into a mathematical expression.

2. The intelligent optimization method for cogeneration based on biomass energy conversion data according to claim 1, characterized in that, The process of collecting and preprocessing key data from the entire biomass energy conversion process and cogeneration system to obtain a standard feature dataset specifically includes: S1.

1. Determine the data collection range and simultaneously use a distributed sensor network to collect data on moisture content, cellulose content, hemicellulose content, lignin content, calorific value, and ash content to obtain biomass raw material characteristic data; S1.

2. Collect pretreatment crushing particle size, pretreatment temperature, pretreatment time, gasification temperature, gasification pressure, equivalence ratio, and residence time to obtain the conversion process parameters; S1.

3. Collect data on CO content, H2 content, CH4 content, calorific value of fuel gas, ash and slag emissions, and NO content in the gasification gas components. x From the concentrations of SO2 and other concentrations, data on the conversion products were obtained. S1.

4. Obtain power generation, heat supply, turbine power, heat exchanger efficiency, thermal storage equipment status, and energy storage equipment status through data acquisition terminals to obtain the operating data of the cogeneration system; S1.

5. Obtain real-time electrical load, real-time heat load, and ambient temperature to determine environmental and demand data; S1.

6. Perform data cleaning on the biomass raw material characteristic data, conversion process parameters and conversion product data collected above, as well as the cogeneration system operation data, environmental and demand data obtained, including missing value filling and outlier removal; S1.

7. Standardize the biomass raw material characteristic data, conversion process parameters, conversion product data, cogeneration system operation data, and environmental and demand data after data cleaning, and use the min-max standardization method to convert parameters of different dimensions to the [0, 1] interval; S1.

8. Perform feature engineering on the standardized biomass feedstock characteristic data, conversion process parameters, conversion product data, cogeneration system operation data, and environmental and demand data to obtain the Pearson correlation coefficient between each characteristic and the overall cogeneration efficiency. S1.

9. Extract the key features with an absolute value of Pearson correlation coefficient greater than 0.6 from the standardized biomass feedstock characteristic data, conversion process parameters, conversion product data, cogeneration system operation data, and environmental and demand data to obtain the initial feature dataset, including moisture content, gasification temperature, equivalence ratio, gas calorific value, real-time electrical load, and real-time heat load. S1.

10. Time-align the temporal data in the initial feature dataset to construct a standard feature dataset.

3. The intelligent optimization method for cogeneration based on biomass energy conversion data according to claim 1, characterized in that, The method involves optimizing parameters based on a mathematical model of a combined heat and power system, a multi-objective optimization objective function, and corresponding constraints, and outputting a Pareto optimal parameter solution set. Specifically, this includes: S4.

1. Select NSGA-II as the intelligent optimization algorithm, and set the population size, maximum number of iterations, crossover probability, and mutation probability; S4.

2. Define the second optimization variable, including the biomass feed rate. Gasifier equivalence ratio Thermoelectric distribution ratio and the charging and discharging power of thermal storage equipment ; S4.

3. Initialize the population by randomly generating initial parameter combinations within the feasible region of the variables. Each individual is represented as: ； S4.

4. Input the individual parameters into the mathematical model of the cogeneration system and calculate the values ​​of each objective function: ； S4.

5. Perform non-dominated ranking, divide individuals into different non-dominated layers according to the multi-objective function value, and calculate the crowding distance to assess individual diversity; S4.

6. Use the tournament selection operator to select parent individuals, and generate child individuals by simulating the binary crossover operator. The formula is: ; In the formula, , For parent parameters, Cross factor; S4.

7. Perform a polynomial mutation operation on the generated offspring individuals, using the following formula: In the formula, As a variable factor, , These are the maximum and minimum values ​​of the variable, respectively. S4.

8. Merge the parent and offspring populations, re-perform the non-dominated sorting and crowding calculation, retain the best individual for the next iteration, until the maximum number of iterations is reached, and output the final Pareto optimal parameter solution set.

4. The intelligent optimization method for cogeneration based on biomass energy conversion data according to claim 1, characterized in that, The establishment of a dynamic iterative optimization mechanism, which updates the model based on real-time acquired new data and triggers dynamic optimization, enables continuous optimized operation of the cogeneration system. Specifically, this includes: S6.

1. Deploy real-time data acquisition terminals to collect real-time key data of biomass raw materials once per hour and store them in a real-time database; S6.

2. Set the model update trigger conditions to trigger the model update process; S6.

3. Use the incremental SVM algorithm to update the data-driven model, training only with new data, and updating the weights using the following formula: In the formula, The original weights, For learning rate, For the actual output, To predict the output, Input features; S6.

4. Execute dynamic optimization trigger judgment, check whether the current parameters deviate from the optimal solution, if so, start the NSGA-II algorithm for incremental optimization; S6.

5. Dynamically adjust and optimize target weights, increasing the weight of efficiency targets during peak electricity consumption periods and increasing the weight of environmental protection targets during environmentally sensitive periods. The weight adjustment formula is as follows: ; In the formula Weights for efficiency objectives Weighting for environmental protection goals, For real-time electrical load, This is the maximum electrical load; S6.

6. Send the new optimized parameters to the control system and achieve real-time adjustment through the PID controller; S6.

7. Record the changes in performance indicators before and after optimization, including efficiency improvement, reduction in pollutant emissions, and reduction in output fluctuation, and generate an optimization effect report; S6.

8. Based on the optimization effect report, evaluate the dynamic iteration mechanism and calculate the optimization stability index: In the formula, For the standard deviation of efficiency fluctuation, This represents the average efficiency.

5. A smart optimization system for cogeneration based on biomass energy conversion data, used to implement the optimization method as described in any one of claims 1-4, characterized in that, include: The data acquisition module is used to collect key data from the entire biomass energy conversion process and also to collect key data from the cogeneration system. The main control module is used to receive data transmitted by the acquisition module, to preprocess key data of the entire biomass energy conversion process and cogeneration system, to obtain standard feature datasets, to construct a mathematical model of cogeneration system that integrates mechanism model and data-driven model based on the standard feature datasets, and to determine multi-objective optimization objective function and corresponding constraints according to the operation requirements of cogeneration system. The optimization module is used to solve for parameters based on the mathematical model of the cogeneration system, the multi-objective optimization objective function and the corresponding constraints, and output the Pareto optimal parameter solution set. It is used to verify the Pareto optimal parameter solution set offline by combining simulation verification and experimental verification, and to establish a dynamic iterative optimization mechanism to update the model based on the new data collected in real time and trigger dynamic optimization to realize the continuous optimized operation of the cogeneration system. The display module is used to present the process and results of intelligent optimization of cogeneration to the user.

6. The intelligent optimization system for cogeneration based on biomass energy conversion data according to claim 5, characterized in that, The acquisition module includes: The first acquisition unit is used to collect key data from the entire biomass energy conversion process. The second acquisition unit is used to acquire key data from the combined heat and power system.

7. The intelligent optimization system for cogeneration based on biomass energy conversion data according to claim 5, characterized in that, The main control module includes: The data receiving unit is used to receive data transmitted by the acquisition module; The data processing unit is used to preprocess key data of the entire biomass energy conversion process and cogeneration system to obtain a standard feature dataset. The model building unit is used to build a mathematical model of a cogeneration system that integrates a mechanism model and a data-driven model based on a standard feature dataset. The model constraint unit is used to determine the multi-objective optimization objective function and corresponding constraint conditions based on the operational requirements of the cogeneration system.

8. The intelligent optimization system for cogeneration based on biomass energy conversion data according to claim 5, characterized in that, The optimization module includes: The optimization solution unit is used to perform parameter optimization based on the mathematical model of the cogeneration system, the multi-objective optimization objective function and the corresponding constraints, and output the Pareto optimal parameter solution set. An offline verification unit is used to perform offline verification of the Pareto optimal parameter solution set by combining simulation verification and experimental verification. The dynamic optimization unit is used to establish a dynamic iterative optimization mechanism, update the model based on real-time acquired new data and trigger dynamic optimization to achieve continuous optimized operation of the cogeneration system.

Images