Intelligent prediction and scheduling management system based on hazardous waste incineration waste heat recovery power generation

Abstract

The invention discloses an intelligent prediction and scheduling management system based on hazardous waste incineration waste heat recovery power generation, relates to the technical field of hazardous waste incineration waste heat recovery power generation, and aims to monitor key parameters such as temperature, pressure, flow and flue gas components in the hazardous waste incineration process in real time, and determine the hazardous waste incineration waste heat recovery power generation through data processing and feature extraction. And the hazardous waste incineration heat value, the waste heat boiler steam production amount and the power generation amount are accurately predicted. Meanwhile, the system combines power grid load change and price fluctuation, realizes intelligent regulation and control of equipment operation and maximization of waste heat recovery efficiency and generating capacity through an optimization scheduling algorithm, can improve energy recovery efficiency to the maximum extent, reduces energy waste, improves economic benefits and environmental sustainability of the system, and has wide application prospects. The power generation benefit can be improved in the field of hazardous waste incineration, and reference can be provided for other industrial waste heat recovery power generation systems.

Description

Technical Field

[0001] The present invention relates to the technical field of waste incineration waste heat recovery power generation, and particularly relates to an intelligent prediction and scheduling management system based on waste incineration waste heat recovery power generation. Background Art

[0002] With the continuous advancement of the global industrialization process, waste treatment and energy recovery have become important issues that need to be urgently solved in the modern industrial environment. Incineration treatment of hazardous waste (hazardous waste) is a relatively common waste disposal method at present. Especially when dealing with hazardous waste with high pollution and complex composition, incineration treatment can not only effectively reduce the volume of waste, but also realize the reuse of energy through waste heat recovery. However, traditional hazardous waste incineration systems face problems such as low waste heat utilization efficiency, serious energy waste, and unstable power generation. There is an urgent need for a more efficient and intelligent way to optimize their operation. Summary of the Invention

[0003] The object of the present invention is to solve the above-mentioned problems, and provide an intelligent prediction and scheduling management system based on waste incineration waste heat recovery power generation.

[0004] The intelligent prediction and scheduling management system based on waste incineration waste heat recovery power generation of the present invention includes:

[0005] Data acquisition module: Real-time acquisition of the operation data of the hazardous waste incinerator, waste heat boiler, and steam turbine equipment, including temperature, pressure, flow rate, and flue gas composition;

[0006] Data processing module: Cleaning and preprocessing the acquired data, and obtaining various types of characteristic data for predicting the calorific value of hazardous waste incineration, the steam production of the waste heat boiler, and the power generation through feature extraction;

[0007] Model prediction module: Using machine learning algorithms to construct a calorific value prediction model for hazardous waste incineration, a steam production prediction model for the waste heat boiler, and a power generation prediction model;

[0008] Optimization scheduling module: According to the prediction results of the calorific value of hazardous waste incineration, the steam production of the waste heat boiler, and the power generation, combined with the change trend of the grid load and the degree of electricity price fluctuation, calculate the optimization scheduling coefficient, and optimize the scheduling of the hazardous waste incineration amount, the operation parameters of the waste heat boiler, and the output of the generator set according to the optimization scheduling coefficient.

[0009] Optionally, the real-time acquisition of the operation data of the hazardous waste incinerator, waste heat boiler, and steam turbine equipment, including temperature, pressure, flow rate, and flue gas composition, is specifically as follows:

[0010] By installing high-precision sensors on the hazardous waste incinerator, waste heat boiler, and steam turbine equipment for real-time monitoring of various operating data, the high-precision sensors include temperature sensors, pressure sensors, flow meters, and flue gas analyzers;

[0011] Ensure that the high-precision sensors are effectively connected to the data acquisition module and data transmission is carried out with the data acquisition module through the industrial Ethernet.

[0012] Optionally, clean and preprocess the collected data, and obtain various types of feature data for predicting the calorific value of hazardous waste incineration, the steam production of the waste heat boiler, and the power generation through feature extraction, as follows:

[0013] The data cleaning includes missing value processing, outlier detection, duplicate data removal, data synchronization and time series calibration;

[0014] The preprocessing includes data standardization and normalization, data smoothing and noise reduction;

[0015] The various types of feature data for predicting the calorific value of hazardous waste incineration, the steam production of the waste heat boiler, and the power generation include the mean, variance, maximum value, minimum value, skewness, kurtosis, frequency components, and lag variables of each operating data.

[0016] Optionally, use machine learning algorithms to construct a calorific value prediction model for hazardous waste incineration, a steam production prediction model for the waste heat boiler, and a power generation prediction model, as follows:

[0017] Construct time series feature data in chronological order with the various types of feature data for predicting the calorific value of hazardous waste incineration, the steam production of the waste heat boiler, and the power generation;

[0018] The calorific value prediction model for hazardous waste incineration, the steam production prediction model for the waste heat boiler, and the power generation prediction model are constructed based on a convolutional neural network, as follows:

[0019] Take the feature data of each time step as channels to construct a two-dimensional matrix. Use a sliding window to divide the time series feature data into multiple subsequences, and each subsequence is used as a sample to form a sample set; divide the sample set into a training set and a test set, with 70% as the training set and 30% as the test set;

[0020] The convolutional neural network includes a convolutional layer, a pooling layer, and a fully connected layer. Based on the three-layer structure of the convolutional layer, pooling layer, and fully connected layer, construct a calorific value prediction model for hazardous waste incineration, a steam production prediction model for the waste heat boiler, and a power generation prediction model respectively;

[0021] Use the training set to train the calorific value prediction model for hazardous waste incineration, the steam production prediction model for the waste heat boiler, and the power generation prediction model, and use the test set to evaluate the prediction accuracy of the model;

[0022] Prediction of hazardous waste incineration calorific value: Input various characteristics of the temperature, pressure, and flow rate of the hazardous waste incinerator into the hazardous waste incineration calorific value prediction model, and output the hazardous waste incineration calorific value;

[0023] Prediction of steam production of waste heat boiler: Input various characteristics of the temperature, pressure, and flue gas composition of the waste heat boiler into the waste heat boiler steam production prediction model, and output the steam production of the waste heat boiler;

[0024] Prediction of power generation: Input various characteristics of the temperature, pressure, and flow rate of the steam turbine into the power generation prediction model, and output the power generation.

[0025] Optionally, the change trend of the grid load is quantified by the grid load change trend coefficient. The acquisition of the grid load change trend coefficient includes:

[0026] Collect the actual load values of the grid at each time point and record the time stamps of the load value data;

[0027] Calculate the grid load value change rate between adjacent time points. The calculation formula is where ΔP t is the grid load value change rate, P t+1 is the actual load value at the (t + 1)-th time point, and P t is the actual load value at the t-th time point;

[0028] Calculate the trend direction coefficient. The calculation formula is where Tdiu is the trend direction coefficient, t = 1, 2, 3, 4, ……, T, and T is a positive integer;

[0029] Calculate the grid load change trend coefficient dwq. The calculation formula is where ΔPmax is the maximum value of the grid load value change rate, and ΔPmin is the minimum value of the grid load value change rate.

[0030] Optionally, the degree of electricity price fluctuation is quantified by the electricity price fluctuation coefficient. The acquisition of the electricity price fluctuation coefficient includes:

[0031] Obtain electricity price data and calculate the electricity price change rate using logarithmic difference. The calculation formula is Dj i = ln(d i + 1) - ln(d i+1 + 1), where Dj i represents the electricity price change rate, d i+1 represents the electricity price at the (i + 1)-th sampling point, and d i represents the electricity price at the i-th sampling point;

[0032] Calculate the mean value of the electricity price change rate. The calculation formula is: where is the mean value of the electricity price change rate, i = 1, 2, 3, 4, ……, N, and N is a positive integer;

[0033] Calculate the standard deviation of the electricity price change rate. The calculation formula is: where oqk is the standard deviation of the electricity price change rate;

[0034] Calculate the electricity price fluctuation coefficient dyb. The calculation formula is

[0035] Optionally, according to the predicted results of the hazardous waste incineration calorific value, the steam production of the waste heat boiler, and the power generation, combined with the grid load change trend and the electricity price fluctuation degree, calculating the optimal scheduling coefficient includes:

[0036]

[0037] In the formula, yhx is the optimal scheduling coefficient, dwq, dyb, wf, yr, fd are the grid load change trend coefficient, electricity price fluctuation coefficient, hazardous waste incineration calorific value, steam production of the waste heat boiler, and power generation respectively, and a1, a2, a3, a4, a5 are the preset proportionality coefficients of dwq, dyb, wf, yr, fd respectively, and a1, a2, a3 are all greater than 0.

[0038] Optionally, compare the optimal scheduling coefficient with the preset optimal scheduling coefficient threshold, and perform optimal scheduling on the hazardous waste incineration amount, the operating parameters of the waste heat boiler, and the output of the generator set, specifically as follows:

[0039] If the optimal scheduling coefficient is not less than the preset optimal scheduling coefficient threshold, generate a scheduling optimization demand signal;

[0040] If the optimal scheduling coefficient is less than the preset optimal scheduling coefficient threshold, do not generate a scheduling optimization demand signal.

[0041] The beneficial effects of the present invention:

[0042] The present invention proposes an intelligent prediction and scheduling management system for waste incineration waste heat recovery power generation. By integrating data acquisition, data processing, model prediction, and optimization scheduling modules, it can achieve efficient prediction of waste incineration calorific value, waste heat boiler steam production, and power generation, and precisely control the system operation through optimization scheduling, thereby achieving the goals of improving energy utilization efficiency, reducing costs, enhancing system stability, and environmental friendliness. By real-time collecting various operation data and using machine learning algorithms for data analysis and prediction, it can dynamically adjust the operation parameters of each device to ensure the optimal operation of the entire system under different loads and electricity price fluctuations. Through optimization scheduling, the system can reduce unnecessary energy consumption and emissions while ensuring power generation efficiency and safety, and improve the recovery utilization rate of waste incineration calorific value. In addition, combined with the grid load change trend and electricity price fluctuation degree, the system can automatically adjust the output of the generator set and the operation parameters of the waste heat boiler, not only improving economic benefits but also enhancing the grid's load adaptability and avoiding unnecessary energy waste and excessive fluctuations. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The following further describes the present invention with reference to the accompanying drawings.

[0044] Figure 1 It is a framework diagram of an intelligent prediction and scheduling management system for waste incineration waste heat recovery power generation. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0045] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0046] Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0047] The embodiments of the present invention provide an intelligent prediction and scheduling management system for waste incineration waste heat recovery power generation. Refer to Figure 1 , Figure 1 which is a framework diagram of an intelligent prediction and scheduling management system for waste incineration waste heat recovery power generation provided by the embodiments of the present invention. The system includes:

[0048] Data acquisition module: Real-time collect the operation data of waste incinerators, waste heat boilers, and steam turbine equipment, including temperature, pressure, flow rate, and flue gas composition;

[0049] Data processing module: Clean and preprocess the collected data, and obtain various types of feature data for predicting the calorific value of hazardous waste incineration, the steam production of waste heat boilers, and the power generation through feature extraction;

[0050] Model prediction module: Use machine learning algorithms to construct a calorific value prediction model for hazardous waste incineration, a steam production prediction model for waste heat boilers, and a power generation prediction model;

[0051] Optimized scheduling module: According to the prediction results of the calorific value of hazardous waste incineration, the steam production of waste heat boilers, and the power generation, calculate the optimized scheduling coefficient in combination with the change trend of the grid load and the degree of electricity price fluctuation, and optimize the scheduling of the hazardous waste incineration volume, the operating parameters of waste heat boilers, and the output of generator sets according to the optimized scheduling coefficient.

[0052] Based on the intelligent prediction and scheduling management system for waste heat recovery power generation from hazardous waste incineration provided by the embodiments of the present invention, in the above manner, key parameters such as temperature, pressure, flow rate, and flue gas composition during the hazardous waste incineration process are monitored in real time, and through data processing and feature extraction, the calorific value of hazardous waste incineration, the steam production of waste heat boilers, and the power generation are accurately predicted. At the same time, the system combines the changes in grid load and price fluctuations, and realizes the intelligent control of equipment operation through an optimized scheduling algorithm, maximizing the waste heat recovery efficiency and power generation, being able to maximize the energy recovery efficiency, reduce energy waste, and at the same time improve the economic benefits and environmental sustainability of the system. It can not only improve the power generation efficiency in the field of hazardous waste incineration, but also provide a reference for other industrial waste heat recovery power generation systems.

[0053] In one embodiment, the operating data of hazardous waste incinerators, waste heat boilers, and steam turbines are collected in real time, including temperature, pressure, flow rate, and flue gas composition, specifically as follows:

[0054] By installing high-precision sensors on hazardous waste incinerators, waste heat boilers, and steam turbines to monitor various operating data in real time, the high-precision sensors include temperature sensors, pressure sensors, flow meters, and flue gas analyzers;

[0055] Ensure that the high-precision sensors are effectively connected to the data acquisition module and transmit data to the data acquisition module through industrial Ethernet;

[0056] It should be noted that before data transmission, edge computing or local caching mechanisms can be used to ensure that the industrial Ethernet temporarily stores data on the premise of stability, avoiding data loss caused by network latency or transmission failure.

[0057] In one embodiment, the collected data is cleaned, preprocessed, and various types of feature data for predicting the calorific value of hazardous waste incineration, the steam production of waste heat boilers, and the power generation are obtained through feature extraction, specifically as follows:

[0058] The data cleaning includes missing value processing, outlier detection, duplicate data removal, data synchronization and time series calibration;

[0059] It should be noted that for detecting and processing missing values in data, common methods include: filling missing values with the mean, median or the previous data point; for time series data, linear interpolation or more complex interpolation methods can be used; for specific data sets, the KNN (K-Nearest Neighbor) interpolation method or multiple imputation method can be used; Outlier detection: Using statistical methods (such as Z-score, IQR) or machine learning algorithms (such as Isolation Forest) to detect and remove outliers. For incorrect sensor data, reasonable replacement values can be adopted or directly discarded; Duplicate data removal: Checking and removing duplicate records in the data to ensure the uniqueness of each data point; Data synchronization and time series calibration: Synchronizing the time of data collected by different devices and sensors to avoid errors caused by inconsistent time series and ensure the temporal consistency of the data, especially in scenarios where multiple devices are running in parallel.

[0060] The preprocessing includes data standardization and normalization, data smoothing and noise reduction;

[0061] It should be noted that standardization: Transforming the data with a mean of 0 and a standard deviation of 1, which is applicable to most machine learning algorithms (such as SVM, KNN, PCA, etc.); Normalization: Scaling the data proportionally to a specified range (such as [0,1]), which is applicable to algorithms sensitive to data scale such as neural networks and gradient descent; Data smoothing and noise reduction: Removing high-frequency noise in the data through smoothing techniques (such as moving average, exponential weighted smoothing, etc.), especially in cases where sensor noise is large.

[0062] The multi-class feature data for predicting the calorific value of hazardous waste incineration, the steam production of the waste heat boiler, and the power generation includes the mean, variance, maximum value, minimum value, skewness, kurtosis, frequency components, and lag variables of each type of operation data;

[0063] It should be noted that the mean is the arithmetic average of all data points, representing the central position of the data. In the prediction of variables such as the calorific value of hazardous waste incineration, the steam production of the boiler, and the power generation, the mean can reflect the overall trend of the system operation. For example, the mean temperature of the incinerator can be used to estimate its combustion efficiency, while the mean steam production of the boiler can reflect the stability of the overall waste heat utilization. The variance is the average of the squared differences between the data points and the mean, measuring the degree of data fluctuation. The variance can reflect the stability of the system operation. A higher variance usually means that the equipment operation fluctuates greatly, which may lead to energy efficiency loss or system failure. For hazardous waste incinerators and waste heat boilers, the variance can help judge the volatility of the calorific value and steam production, thus predicting possible efficiency reduction or failure risks. The maximum value is the largest numerical value in the data set, and the minimum value is the smallest numerical value in the data set. The maximum and minimum values can be used to evaluate the extreme state of the equipment at a certain moment. For example, the maximum and minimum values of the temperature can help monitor the thermal control range of the incinerator to ensure that the combustion process does not experience overheating or overcooling, thereby preventing energy efficiency loss or equipment damage. Skewness measures the degree of skewness of the data distribution, describing the symmetry of the data distribution. Positive skewness indicates that the right side of the data distribution is longer, while negative skewness indicates that the left side is longer. Skewness helps judge the asymmetry of the data distribution, which is very important for understanding the combustion instability in the incineration process or the equipment fluctuations in the waste heat recovery system. For example, the skewness of the incineration temperature may reflect whether there is unevenness in the combustion process, affecting the overall thermal efficiency of the system. Kurtosis describes the sharpness of the data distribution, measuring the degree of data concentration. A higher kurtosis indicates a stronger concentration of the data, while a lower kurtosis indicates a more flat data distribution. Kurtosis can reveal extreme fluctuations in the data. For the hazardous waste incineration system, if the kurtosis of the temperature data is high, it may mean that there are abnormal fluctuations, which may be signs of equipment failure or incomplete combustion, and further diagnosis and prediction are required. The frequency components analyze the frequency distribution in the time series data through Fourier transform (FFT) or other frequency domain analysis methods. The frequency components help analyze the periodic fluctuations in the data. For example, the steam production in the waste heat boiler may have periodic fluctuations, and these periodic patterns can be extracted through the frequency components to help predict the stability and efficiency of the boiler under different loads. The lag variable refers to the relationship between the data at the current moment and the data at a certain past moment in time series analysis. For example, the data lagged by 1 moment is the value of the previous time point of the current data. The lag variable is crucial in time series prediction because the operating states of many equipment are affected by historical data. For example, the temperature change of the incinerator may depend on the operating state in the past period of time, and the lag variable can help the model capture this time series dependence and improve the prediction accuracy.

[0064] In one embodiment, machine learning algorithms are used to construct a hazardous waste incineration calorific value prediction model, a waste heat boiler steam production prediction model, and an electricity generation prediction model, as follows:

[0065] Construct time series feature data in chronological order for various types of feature data used to predict the calorific value of hazardous waste incineration, the steam production of waste heat boilers, and electricity generation;

[0066] The above-mentioned hazardous waste incineration calorific value prediction model, waste heat boiler steam production prediction model, and electricity generation prediction model are constructed based on a convolutional neural network, as follows:

[0067] Take the feature data of each time step as channels to construct a two-dimensional matrix. Use a sliding window to divide the time series feature data into multiple subsequences, and each subsequence is used as a sample to form a sample set; divide the sample set into a training set and a test set, with 70% as the training set and 30% as the test set;

[0068] The convolutional neural network includes a convolutional layer, a pooling layer, and a fully connected layer. Based on the three-layer structure of the convolutional layer, pooling layer, and fully connected layer, construct a hazardous waste incineration calorific value prediction model, a waste heat boiler steam production prediction model, and an electricity generation prediction model respectively;

[0069] Use the training set to train the hazardous waste incineration calorific value prediction model, the waste heat boiler steam production prediction model, and the electricity generation prediction model, and use the test set to evaluate the prediction accuracy of the model;

[0070] Hazardous waste incineration calorific value prediction: Input various types of features such as the temperature, pressure, and flow rate of the hazardous waste incinerator into the hazardous waste incineration calorific value prediction model to output the hazardous waste incineration calorific value;

[0071] Waste heat boiler steam production prediction: Input various types of features such as the temperature, pressure, and flue gas composition of the waste heat boiler into the waste heat boiler steam production prediction model to output the waste heat boiler steam production;

[0072] Electricity generation prediction: Input various types of features such as the temperature, pressure, and flow rate of the steam turbine into the electricity generation prediction model to output the electricity generation.

[0073] In one embodiment, the changing trend of the grid load is quantified by a grid load change trend coefficient, and the acquisition of the grid load change trend coefficient includes:

[0074] Collect the actual load values of the grid at each time point and record the timestamps of the load value data;

[0075] Calculate the grid load value change rate between adjacent time points. The calculation formula is where ΔP t is the grid load value change rate, and P t+1 is the actual load value at the (t + 1)-th time point, and P tis the actual load value at the t-th time point;

[0076] Calculate the trend direction coefficient, and the calculation formula is where Tdiu is the trend direction coefficient, t = 1, 2, 3, 4, ……, T, and T is a positive integer;

[0077] Calculate the power grid load change trend coefficient dwq, and the calculation formula is where ΔPmax is the maximum value of the power grid load value change rate, and ΔPmin is the minimum value of the power grid load value change rate.

[0078] In one embodiment, the degree of electricity price fluctuation is quantified by an electricity price fluctuation coefficient, and the acquisition of the electricity price fluctuation coefficient includes:

[0079] Obtain electricity price data and use logarithmic difference to calculate the electricity price change rate. The calculation formula is Dj i = ln(d i + 1) - ln(d i+1 + 1), where Dj i represents the electricity price change rate, d i+1 represents the electricity price at the (i + 1)-th sampling point, and d i represents the electricity price at the i-th sampling point;

[0080] Calculate the mean value of the electricity price change rate. The calculation formula is: where is the mean value of the electricity price change rate, i = 1, 2, 3, 4, ……, N, and N is a positive integer;

[0081] Calculate the standard deviation of the electricity price change rate. The calculation formula is: where oqk is the standard deviation of the electricity price change rate;

[0082] Calculate the electricity price fluctuation coefficient dyb. The calculation formula is

[0083] In one embodiment, according to the predicted results of the hazardous waste incineration calorific value, the steam production of the waste heat boiler, and the power generation, combining the power grid load change trend and the degree of electricity price fluctuation to calculate the optimal scheduling coefficient includes:

[0084] yhx = (a1 × dwq + a2 × dyb) × e (a3×wf+a4×yr+a5×fd)

[0085] In the formula, yhx is the optimal scheduling coefficient, dwq, dyb, wf, yr, and fd are the power grid load change trend coefficient, the electricity price fluctuation coefficient, the hazardous waste incineration calorific value, the steam production of the waste heat boiler, and the power generation respectively, and a1, a2, a3, a4, and a5 are the preset proportional coefficients of dwq, dyb, wf, yr, and fd respectively, and a1, a2, and a3 are all greater than 0.

[0086] It should be noted that a1, a2, a3, a4, and a5 are set by professionals according to the actual situation. Generally, the sum of a1, a2, a3, a4, and a5 is 1. For example, a1, a2, a3, a4, and a5 can be 0.2, 0.3, 0.3, 0.1, and 0.1 respectively, or they can be other numbers, and there is no specific limitation;

[0087] In one embodiment, the optimized scheduling coefficient is compared with the preset optimized scheduling coefficient threshold, and the optimized scheduling of the hazardous waste incineration amount, the operating parameters of the waste heat boiler, and the output of the generator set is carried out as follows:

[0088] If the optimized scheduling coefficient is not less than the preset optimized scheduling coefficient threshold, the demand for optimizing the scheduling of the hazardous waste incineration amount, the operating parameters of the waste heat boiler, and the output of the generator set is relatively high, and a scheduling optimization demand signal is generated;

[0089] If the optimized scheduling coefficient is less than the preset optimized scheduling coefficient threshold, the demand for optimizing the scheduling of the hazardous waste incineration amount, the operating parameters of the waste heat boiler, and the output of the generator set is relatively low, and no scheduling optimization demand signal is generated.

[0090] In one implementation manner, the above-mentioned preset optimized scheduling coefficient threshold is set by professionals according to the actual situation, and the details are not described herein again.

[0091] In one implementation manner, by integrating data acquisition, data processing, model prediction, and optimized scheduling modules, it is possible to achieve efficient prediction of the calorific value of hazardous waste incineration, the steam production of the waste heat boiler, and the power generation, and to precisely control the system operation through optimized scheduling, so as to achieve the goals of improving energy utilization efficiency, reducing costs, enhancing system stability, and environmental friendliness. By real-time collecting various operation data and using machine learning algorithms for data analysis and prediction, the operation parameters of each device can be dynamically adjusted to ensure the optimal operation of the entire system under different loads and electricity price fluctuations. Through optimized scheduling, the system can reduce unnecessary energy consumption and emissions while ensuring power generation efficiency and safety, and improve the recovery utilization rate of the calorific value of hazardous waste incineration. In addition, by combining the grid load change trend and the degree of electricity price fluctuation, the system can automatically adjust the output of the generator set and the operation parameters of the waste heat boiler, which not only improves economic benefits but also enhances the load adaptability of the grid, avoiding unnecessary energy waste and excessive fluctuations.

[0092] The above has described a detailed description of an embodiment of the present invention, but the content described is only a preferred embodiment of the present invention and cannot be artificially used to limit the implementation scope of the present invention. All equivalent changes and improvements made within the scope of the application of the present invention should still fall within the scope covered by the patent of the present invention.

Claims

1. An intelligent prediction and dispatching management system based on waste heat recovery from hazardous waste incineration for power generation, characterized in that: The system comprises: Data acquisition module: real-time collection of operating data of hazardous waste incinerators, waste heat boilers, and steam turbine equipment, including temperature, pressure, flow, and flue gas composition; Data processing module: cleans and preprocesses the collected data, and obtains multiple types of feature data for predicting the calorific value of hazardous waste incineration, steam production of waste heat boilers, and power generation through feature extraction; Model prediction module: Use machine learning algorithms to build hazardous waste incineration calorific value prediction models, waste heat boiler steam production prediction models, and power generation prediction models; Optimization and dispatching module: Based on the predicted results of hazardous waste incineration calorific value, waste heat boiler steam production, and power generation, combined with the grid load change trend and electricity price fluctuation degree, the optimization dispatching coefficient is calculated, and the hazardous waste incineration amount, waste heat boiler operating parameters, and generator set output are optimized according to the optimization dispatching coefficient.

2. The intelligent prediction and dispatching management system based on waste heat recovery from hazardous waste incineration for power generation according to claim 1 is characterized in that: Real-time collection of operating data of hazardous waste incinerators, waste heat boilers, and steam turbine equipment, including temperature, pressure, flow, and flue gas composition, as follows: By installing high-precision sensors on hazardous waste incinerators, waste heat boilers, and steam turbine equipment to monitor various operating data in real time, the high-precision sensors include temperature sensors, pressure sensors, flow meters, and flue gas analyzers; Ensure that the high-precision sensor is effectively connected to the data acquisition module and transmit data with the data acquisition module through industrial Ethernet.

3. The intelligent prediction and dispatching management system based on hazardous waste incineration waste heat recovery power generation according to claim 2 is characterized in that: The collected data is cleaned and preprocessed, and multiple types of feature data for predicting the calorific value of hazardous waste incineration, steam production of waste heat boilers, and power generation are obtained through feature extraction, as follows: The data cleaning includes missing value processing, outlier detection, duplicate data removal, data synchronization and time series calibration; The preprocessing includes data standardization and normalization, data smoothing and noise reduction; The multiple types of characteristic data used to predict the calorific value of hazardous waste incineration, steam production of waste heat boilers, and power generation include the mean, variance, maximum value, minimum value, skewness, kurtosis, frequency component, and lag variables of each type of operating data.

4. The intelligent prediction and dispatching management system based on waste heat recovery from hazardous waste incineration for power generation according to claim 1 is characterized in that: The machine learning algorithm is used to build a hazardous waste incineration calorific value prediction model, a waste heat boiler steam production prediction model, and a power generation prediction model, as follows: The multi-category feature data used to predict the calorific value of hazardous waste incineration, steam production of waste heat boilers, and power generation are constructed into time series feature data in chronological order; The hazardous waste incineration calorific value prediction model, waste heat boiler steam production prediction model, and power generation prediction model are constructed based on convolutional neural networks, as follows: The characteristic data of each time step is used as a channel to construct a two-dimensional matrix. The time series characteristic data is divided into multiple subsequences using a sliding window. Each subsequence is used as a sample to form a sample set. The sample set is divided into a training set and a test set, with 70% being the training set and 30% being the test set. The convolutional neural network includes a convolutional layer, a pooling layer, and a fully connected layer. The convolutional neural network based on the three-layer structure of the convolutional layer, the pooling layer, and the fully connected layer respectively constructs a hazardous waste incineration calorific value prediction model, a waste heat boiler steam production prediction model, and a power generation prediction model; Use the training set to train the hazardous waste incineration calorific value prediction model, waste heat boiler steam production prediction model, and power generation prediction model, and use the test set to evaluate the prediction accuracy of the model; Hazardous waste incineration calorific value prediction: Input multiple types of characteristics of hazardous waste incinerator temperature, pressure, and flow into the hazardous waste incineration calorific value prediction model to output the hazardous waste incineration calorific value; Prediction of waste heat boiler steam production: Input the temperature, pressure, and flue gas composition characteristics of the waste heat boiler into the waste heat boiler steam production prediction model, and output the waste heat boiler steam production; Power generation prediction: Input multiple types of features such as temperature, pressure and flow of the steam turbine into the power generation prediction model and output the power generation.

5. The intelligent prediction and dispatching management system based on hazardous waste incineration waste heat recovery power generation according to claim 1 is characterized in that: The grid load variation trend is quantified by the grid load variation trend coefficient. The grid load variation trend coefficient is obtained by: Collect the actual load value of the power grid at each time point and record the timestamp of the load value data; Calculate the rate of change of power grid load value at adjacent time points. The calculation formula is: Where ΔP t is the rate of change of power grid load value, P t+1 is the actual load value at the t+1th time point, P t is the actual load value at the tth time point; Calculate the trend direction coefficient, the calculation formula is Where Tdiu is the trend direction coefficient, t = 1, 2, 3, 4, ..., T, and T is a positive integer; Calculate the power grid load change trend coefficient dwq, the calculation formula is: Among them, ΔPmax is the maximum value of the grid load value change rate, and ΔPmin is the minimum value of the grid load value change rate.

6. The intelligent prediction and dispatching management system based on waste heat recovery from hazardous waste incineration for power generation according to claim 1 is characterized in that: The degree of electricity price fluctuation is quantified by the electricity price fluctuation coefficient, and the acquisition of the electricity price fluctuation coefficient includes: The electricity price data is obtained by using logarithmic difference to calculate the rate of change of electricity price. The calculation formula is Dj i =ln(d i +1)-ln(d i+1 +1), where Dj i represents the rate of change of electricity price, d i+1 represents the electricity price at the i+1th sampling point, d i represents the electricity price at the i-th sampling point; Calculate the mean of the electricity price change rate using the following formula: in is the mean value of the electricity price change rate, i = 1, 2, 3, 4, ..., N, and N is a positive integer; Calculate the standard deviation of the electricity price change rate using the following formula: Where oqk is the standard deviation of the rate of change of electricity prices; Calculate the electricity price fluctuation coefficient dyb, the calculation formula is:

7. The intelligent prediction and dispatching management system based on waste heat recovery from hazardous waste incineration for power generation according to claim 1 is characterized in that: According to the prediction results of hazardous waste incineration calorific value, waste heat boiler steam production and power generation, combined with the load change trend of the power grid and the degree of electricity price fluctuation, the optimization dispatch coefficient is calculated, including: Where yhx is the optimization dispatch coefficient, dwq, dyb, wf, yr, and fd are the grid load change trend coefficient, electricity price fluctuation coefficient, hazardous waste incineration calorific value, waste heat boiler steam production, and power generation, respectively; a1, a2, a3, a4, and a5 are the preset proportional coefficients of dwq, dyb, wf, yr, and fd, respectively, and a1, a2, and a3 are all greater than 0.

8. The intelligent prediction and dispatching management system based on waste heat recovery from hazardous waste incineration for power generation according to claim 7 is characterized in that: The optimized dispatch coefficient is compared with the preset optimized dispatch coefficient threshold, and the hazardous waste incineration amount, waste heat boiler operating parameters, and generator output are optimized as follows: If the optimization scheduling coefficient is not less than the preset optimization scheduling coefficient threshold, a scheduling optimization demand signal is generated; If the optimization scheduling coefficient is less than the preset optimization scheduling coefficient threshold, no scheduling optimization demand signal is generated.

Images