Deep peak regulation energy saving method based on 20% load

Abstract

The invention discloses a deep peak regulation energy-saving method based on 20% load, and relates to the technical field of deep peak regulation. Comprising the steps of obtaining a real-time load and a target load of a current unit, determining an adjustment instruction of a unit load based on a difference value between the real-time load and the target load, obtaining operation data of the unit and cost data of load adjustment, determining a load adjustment sequence of the unit capable of carrying out load adjustment and an adjustable unit, the method comprises the following steps: determining the load regulation capacity of all adjustable units, distributing a load regulation target according to a load regulation sequence of the adjustable units, regulating the load of the units to a target load, constructing a double-input double-output model for unit coordination control, optimizing and updating the model by adopting a particle swarm optimization algorithm, and setting a fitness function to search an optimal solution. And determining a unit adjusting instruction. Accurate data support can be provided for planning operation of the power plant, unit loss is reduced, and operation efficiency and energy utilization efficiency of the power plant are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of deep peak regulation, and in particular to a deep peak regulation and energy-saving method based on 20% load. Background Art

[0002] Coal-fired power plants are an important part of my country's power industry. Their operating efficiency and energy consumption are directly related to the country's energy security and environmental protection. However, with the continuous growth of electricity demand and the adjustment of energy structure, coal-fired power plants are facing increasing peak-shaving pressure. Existing performance evaluation methods for coal-fired power plants participating in deep peak-shaving are mostly based on the power load output limit of the coal-fired power plant's overall participation in regulation or the timeliness of its participation in regulation. However, this method ignores the differences between different units within the coal-fired power plant and focuses too much on the maximum adjustable capacity of the coal-fired units. Especially when the load is as low as 20%, the adjustment cost is too high in order to achieve the target load, and even affects the safe operation of the coal-fired units, and cannot accurately reflect the actual operating conditions and effects of the units.

[0003] Therefore, providing a deep peak-shaving energy-saving method based on 20% load to solve the difficulties existing in the prior art is an issue that needs to be urgently addressed by those skilled in the art. Summary of the Invention

[0004] In view of this, the present invention provides a deep peak-shaving energy-saving method based on 20% load, which can provide accurate data support for the planning and operation of power plants, reduce unit losses, and improve the operating efficiency and energy utilization efficiency of power plants.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A deep peak-shaving energy-saving method based on 20% load includes the following steps:

[0007] Obtain the real-time load and target load of the current unit, and determine the unit load adjustment instruction based on the difference between the real-time load and the target load;

[0008] Obtain the operating data of the units and the cost data of load regulation, and determine the units that can be load regulated and the load regulation sequence of the adjustable units;

[0009] Determine the load regulation capabilities of all adjustable units, assign load regulation targets according to the load regulation sequence of the adjustable units, and adjust the load of the units to the target load;

[0010] A dual-input and dual-output model for unit coordinated control is constructed, the particle swarm optimization algorithm is used to optimize and update the model, the fitness function is set to search for the optimal solution, and the unit adjustment instructions are determined.

[0011] Optional load adjustment sequences for adjustable units include:

[0012] Identify adjustable coal-fired units and calculate the load adjustment costs and carbon emissions of these adjustable units;

[0013] Obtain real-time load data and historical load data of the adjustable coal-fired unit, determine the stable load mean of the coal-fired unit based on the historical load data, and calculate the difference between the real-time load data and the stable load mean;

[0014] Standardize the load regulation cost, carbon emission data, and difference data, and assign corresponding weights to the standardized load regulation cost, carbon emission data, and difference data;

[0015] The priorities of the adjustable coal-fired units are calculated based on the weights of the standardized load regulation cost, carbon emission data and difference data, and the priorities of the adjustable coal-fired units are arranged from large to small to obtain the load regulation sequence of the adjustable units.

[0016] Optionally, in the dual-input dual-output model, the output is set to the unit output power N and the unit pressure P T , the input is the valve opening μ T and coal feed rate μ B , the corresponding model expression is:

[0017]

[0018] Among them, W NT (s) is the valve opening μ T Transfer function of the unit output power N, W NB (s) is the coal feeding amount μ B Transfer function of the unit output power N, W PT (s) is the valve opening μ T With the unit pressure P T The transfer function, W PB (s) is the coal feeding amount μ B With the unit pressure P T The transfer function of .

[0019] Optionally, the unit output power is also obtained based on deep peak-shaving power. The deep peak-shaving power acquisition methods include:

[0020] Obtain historical data, real-time data, and target load data for coal-fired units;

[0021] Perform data preprocessing on historical data, real-time data, and target load data, select a time series prediction model based on the preprocessed historical data for training and prediction, evaluate the performance of the trained model, and obtain the final prediction model;

[0022] The power generation forecast results are obtained based on the prediction model combined with the standardized real-time data and target load data.

[0023] Optionally, the time series prediction model includes: linear regression model, decision tree model, support vector machine model and neural network model.

[0024] Optional model optimization updates include:

[0025] Initialize the particle swarm algorithm parameters and update the particle speed and position based on the dual-input dual-output model;

[0026] The particle swarm is iteratively updated, and the fitness function and fuzzy control are combined to determine whether the termination conditions are met. If so, the optimal solution is output; if not, the calculation is updated.

[0027] It can be seen from the above technical solution that compared with the existing technology, the present invention provides a deep peak-shaving energy-saving method based on 20% load, which has the following beneficial effects: 1) The present invention takes into account the cost of load regulation, meets the changes in external load demand, and enables various energy resources to be used more reasonably and efficiently; 2) The present invention combines time series prediction models and algorithms to train and evaluate the prediction models, make accurate power generation predictions, and provide accurate data support for the planning and operation of power plants; 3) The present invention uses fuzzy control to adaptively adjust the units, improve the control accuracy of the unit operation and reduce the unit loss caused by the control operation, thereby improving the operating efficiency and energy utilization efficiency of the power plant. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0029] Figure 1 This is a flow chart of a deep peak-shaving energy-saving method based on 20% load disclosed by the present invention. DETAILED DESCRIPTION

[0030] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0031] Reference Figure 1 As shown, the present invention discloses a deep peak-shaving energy-saving method based on 20% load, comprising the following steps:

[0032] Obtain the real-time load and target load of the current unit, and determine the unit load adjustment instruction based on the difference between the real-time load and the target load;

[0033] Obtain the operating data of the units and the cost data of load regulation, and determine the units that can be load regulated and the load regulation sequence of the adjustable units;

[0034] Determine the load regulation capabilities of all adjustable units, assign load regulation targets according to the load regulation sequence of the adjustable units, and adjust the load of the units to the target load;

[0035] A dual-input and dual-output model for unit coordinated control is constructed, the particle swarm optimization algorithm is used to optimize and update the model, the fitness function is set to search for the optimal solution, and the unit adjustment instructions are determined.

[0036] Furthermore, the adjustment instruction includes determining the real-time load of the coal-fired machine, judging the real-time load and the target load, and if the real-time load exceeds the target load, setting the adjustment instruction to reduce the load; if the real-time load is less than the target load, setting the adjustment instruction to increase the load.

[0037] Furthermore, determining the adjustable load units includes evaluating the unit adjustment safety factor, including:

[0038] Determine the safety indicators and corresponding test values of adjustable units;

[0039] Assign weights to the detection values and obtain evaluation scores based on safety indicators;

[0040] An evaluation threshold is set and coal-fired units with evaluation scores exceeding the threshold are marked as adjustable units.

[0041] Specifically, deep peak regulation will cause losses to coal-fired units and thus pose a greater safety hazard. Therefore, the set safety indicators include: heating surface indicators: the air flow velocity in the furnace is reduced, and the fuel combustion is incomplete, which will cause uneven distribution of fly ash concentration, resulting in increased wear of the heating surface. At the same time, due to changes in combustion conditions, acidic gases such as sulfur dioxide in the flue gas combine with water vapor to form acidic condensate on the low-temperature heating surface, causing low-temperature corrosion; stator winding indicators: the load of the generator changes frequently, and the current and temperature in the stator winding also change accordingly, which accelerates the aging of the stator winding insulation, reduces the insulation performance, and increases the risk of short-circuit failure; rotor winding indicators: the excitation current of the generator may increase, causing the rotor winding to heat up. If the heat dissipation conditions are not good, the rotor winding temperature will be too high, which will affect its insulation performance and mechanical strength, and may even cause the rotor winding to be grounded or short-circuited.

[0042] Furthermore, the load adjustment sequence of the adjustable unit includes:

[0043] Identify adjustable coal-fired units and calculate the load adjustment costs and carbon emissions of these adjustable units;

[0044] Obtain real-time load data and historical load data of the adjustable coal-fired unit, determine the stable load mean of the coal-fired unit based on the historical load data, and calculate the difference between the real-time load data and the stable load mean;

[0045] Standardize the load regulation cost, carbon emission data, and difference data, and assign corresponding weights to the standardized load regulation cost, carbon emission data, and difference data;

[0046] The priorities of the adjustable coal-fired units are calculated based on the weights of the standardized load regulation cost, carbon emission data and difference data, and the priorities of the adjustable coal-fired units are arranged from large to small to obtain the load regulation sequence of the adjustable units.

[0047] Furthermore, in the dual-input dual-output model, the output is set to the unit output power N and the unit pressure P T , the input is the valve opening μ T and coal feed rate μ B , the corresponding model expression is:

[0048]

[0049] Among them, W NT (s) is the valve opening μ T Transfer function of the unit output power N, W NB (s) is the coal feeding amount μ B Transfer function of the unit output power N, W PT (s) is the valve opening μ T and unit pressure P T The transfer function, W PB (s) is the coal feeding amount μ B and unit pressure P T The transfer function of .

[0050] Furthermore, the unit output power is also obtained based on deep peak-shaving power, and the deep peak-shaving power acquisition method includes:

[0051] Obtain historical data, real-time data, and target load data for coal-fired units;

[0052] Perform data preprocessing on historical data, real-time data, and target load data, select a time series prediction model based on the preprocessed historical data for training and prediction, evaluate the performance of the trained model, and obtain the final prediction model;

[0053] The power generation forecast results are obtained based on the prediction model combined with the standardized real-time data and target load data.

[0054] Furthermore, data preprocessing methods include determining the time range and frequency of the data, whether the data is missing or abnormal, and performing data cleaning on the missing and abnormal data.

[0055] Furthermore, time series prediction models include: linear regression model, decision tree model, support vector machine model and neural network model.

[0056] Specifically, in the linear regression model, power generation is used as the dependent variable, and the operating status characteristics of the power generation equipment are used as the independent variable. The operating data of various indicators of the power generation equipment that affect power generation are collected. The data is collected through sensors and detection systems, and the collected data is screened. Through correlation analysis, independent variables with high correlation with power generation are found and included in the model. The regression equation coefficients are obtained through the least squares method. Data is collected through statistical software for regression analysis to obtain the regression equation. The prediction accuracy of the model is evaluated by verifying the goodness of fit.

[0057] In the decision tree model, the equipment operation data is collected and the XGBoost algorithm is used to train and generate a gradient boosting decision tree model. The equipment operation data is used to obtain the power generation data through the decision tree model, with the data of the power generation equipment as the input value and the power generation as the output value;

[0058] The support vector machine model uses power generation equipment information to derive the power generation of the gas turbine. The input includes historical data and power generation equipment information data, and the output is power generation and power generation.

[0059] The neural network model uses the error after the output to estimate the error of the direct predecessor layer of the output layer, and uses the error to estimate the error of the previous layer. If the actual output of the output layer does not match the expected output, the error is distributed to all units in each layer, thereby correcting the weights of the units in each layer. When the output error reaches an acceptable level, the training ends.

[0060] Further model optimization and updates include:

[0061] Initialize the particle swarm algorithm parameters and update the particle speed and position based on the dual-input dual-output model;

[0062] The particle swarm is iteratively updated, and the fitness function and fuzzy control are combined to determine whether the termination conditions are met. If so, the optimal solution is output; if not, the calculation is updated.

[0063] Specifically, the particle swarm algorithm initializes a group of particles and uses swarm cooperation and information sharing to find the optimal solution. Each particle has speed and position attributes, which are adjusted based on individual and global extreme values. To improve efficiency, an inertia weight factor is introduced to adjust the speed. The particle's speed and position are updated based on a specific learning factor and a random number. The fitness function uses a triangular membership function, and the fuzzy controller uses the deviation and the rate of change of the deviation as inputs to output fuzzy control parameters. The fuzzy control parameters obtained after the particle swarm algorithm converges are the optimal solution.

[0064] Furthermore, model optimization also includes ensuring that the grid power obtained by the unit's output power is not less than the predicted power generation.

[0065] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A deep peak-shaving energy-saving method based on 20% load, characterized in that: The following steps are involved: Obtain the real-time load and target load of the current unit, and determine the unit load adjustment instruction based on the difference between the real-time load and the target load; Obtain the operating data of the units and the cost data of load regulation, and determine the units that can be load regulated and the load regulation sequence of the adjustable units; Determine the load regulation capabilities of all adjustable units, assign load regulation targets according to the load regulation sequence of the adjustable units, and adjust the load of the units to the target load; A dual-input and dual-output model for unit coordinated control is constructed, the particle swarm optimization algorithm is used to optimize and update the model, the fitness function is set to search for the optimal solution, and the unit adjustment instructions are determined.

2. The deep peak-shaving energy-saving method based on 20% load according to claim 1 is characterized in that: The load regulation sequence for adjustable units includes: Identify adjustable coal-fired units and calculate the load adjustment costs and carbon emissions of these adjustable units; Obtain real-time load data and historical load data of the adjustable coal-fired unit, determine the stable load mean of the coal-fired unit based on the historical load data, and calculate the difference between the real-time load data and the stable load mean; Standardize the load regulation cost, carbon emission data, and difference data, and assign corresponding weights to the standardized load regulation cost, carbon emission data, and difference data; The priorities of the adjustable coal-fired units are calculated based on the weights of the standardized load regulation cost, carbon emission data and difference data, and the priorities of the adjustable coal-fired units are arranged from large to small to obtain the load regulation sequence of the adjustable units.

3. The deep peak-shaving energy-saving method based on 20% load according to claim 1 is characterized in that: In the dual-input dual-output model, the output is set to the unit output power N and the unit pressure P T , the input is the valve opening μ T and coal feed rate μ B , the corresponding model expression is: Among them, W NT (s) is the valve opening μ T Transfer function of the unit output power N, W NB (s) is the coal feeding amount μ B Transfer function of the unit output power N, W PT (s) is the valve opening μ T and unit pressure P T The transfer function, W PB (s) is the coal feeding amount μ B and unit pressure P T The transfer function of .

4. The deep peak load regulation and energy saving method based on 20% load according to claim 3 is characterized in that: The unit output power is also obtained based on deep peak-shaving power. The deep peak-shaving power acquisition methods include: Obtain historical data, real-time data, and target load data for coal-fired units; Perform data preprocessing on historical data, real-time data, and target load data, select a time series prediction model based on the preprocessed historical data for training and prediction, evaluate the performance of the trained model, and obtain the final prediction model; The power generation forecast results are obtained based on the prediction model combined with the standardized real-time data and target load data.

5. The deep peak load regulation and energy saving method based on 20% load according to claim 4 is characterized in that: Time series prediction models include: linear regression model, decision tree model, support vector machine model and neural network model.

6. The deep peak load regulation and energy saving method based on 20% load according to claim 1 is characterized in that: Model optimization updates include: Initialize the particle swarm algorithm parameters and update the particle speed and position based on the dual-input dual-output model; The particle swarm is iteratively updated, and the fitness function and fuzzy control are combined to determine whether the termination conditions are met. If so, the optimal solution is output; if not, the calculation is updated.

Images