A steel enterprise gas-electricity collaborative optimization scheduling method based on improved genetic algorithm

Abstract

The application discloses a steel enterprise coal gas-power collaborative optimization scheduling method based on an improved genetic algorithm, and belongs to the technical field of energy optimization scheduling of steel enterprises.The application solves the problems of the existing method, such as lack of deep cooperation between coal gas and power, imperfect double-energy storage constraint system, low convergence efficiency of the optimization process and poor engineering feasibility.The application constructs an input quantity system with hourly coal gas production, front-end coal gas consumption, power demand, time-of-use electricity price, initial state of a gas storage tank and a power station, and an output quantity system with hourly power purchase amount, gas consumption of each power station, charging and discharging amount and various types of coal gas emission amount, and constructs an optimization target according to total power purchase cost and coal gas emission amount of the steel enterprise, so as to realize deep cooperation between coal gas and power.The application introduces a coal gas storage tank capacity constraint and a power station power storage constraint to perfect the double-energy storage constraint system of coal gas and power, and improves the genetic algorithm to improve the convergence efficiency and engineering feasibility.The application can be applied to the collaborative optimization scheduling of coal gas and power.

Description

Technical Field

[0001] This invention belongs to the field of energy optimization and scheduling technology for steel enterprises, specifically involving a gas-electricity collaborative optimization scheduling method for steel enterprises based on an improved genetic algorithm. Background Technology

[0002] As a pillar industry of the national economy, the steel industry has a high energy consumption. Blast furnace gas (BFG), coke oven gas (COG), and converter gas (LDG) produced during steel production are three core by-product fuel gases, with annual emissions exceeding 100 billion cubic meters. They are not only important secondary energy sources but also key tools for enterprises to reduce costs, increase efficiency, and reduce carbon emissions.

[0003] However, by-product coal gas has inherent characteristics such as large spatial and temporal fluctuations in production, significant differences in calorific value, and limited storage and regulation capabilities. Meanwhile, there is a strong coupling relationship between the electricity demand and coal gas supply of steel enterprises. Electricity demand is not only enormous in total volume but also characterized by significant peak-valley price differences and uneven demand across different time periods; peak-hour electricity prices sometimes exceed valley-hour prices by more than three times. The intertwined fluctuations on both the supply and demand sides make energy dispatching extremely complex.

[0004] Traditional steel companies often rely on static allocation strategies based on manual experience, resulting in high gas venting rates, high external electricity purchase costs, and poor energy synergy. While existing research has incorporated mathematical programming and intelligent optimization algorithms, significant limitations remain: most studies focus on optimizing production from a single gas source and fail to adequately characterize the coupling relationship between hourly gas fluctuations and peak / valley electricity demand, leading to a lack of deep synergy between gas and electricity; the dual energy storage constraint design for both gas and electricity is inadequate; and traditional algorithms suffer from insufficient convergence efficiency and engineering feasibility in multi-constraint dynamic scheduling scenarios.

[0005] Therefore, there is an urgent need to develop an efficient and optimized scheduling method that can simultaneously consider gas characteristics, peak and off-peak electricity prices, and dual energy storage constraints, so as to achieve dual optimization of energy costs and energy consumption for steel enterprises. Summary of the Invention

[0006] This invention addresses the problems of existing methods, such as lack of deep coordination between gas and electricity, imperfect dual energy storage constraint system, low convergence efficiency of optimization process, and poor engineering feasibility. It proposes a gas-electricity coordinated optimization scheduling method for steel enterprises based on an improved genetic algorithm.

[0007] The technical solution adopted by this invention to solve the above-mentioned technical problems is: a gas-electricity coordinated optimization scheduling method for steel enterprises based on an improved genetic algorithm, the method specifically including the following steps:

[0008] Step 1: Input the time dimension parameters, power plant parameters, gas parameters, power parameters, and genetic algorithm parameters;

[0009] Step 2: Map the hourly electricity purchase and the hourly gas consumption of each power station into continuous scheduling decision variables. Use the peak-valley differentiated pre-allocation mechanism and scheduling decision variables to generate an initial population. Each individual in the initial population is an hourly scheduling scheme corresponding to the entire scheduling cycle.

[0010] Step 3: Calculate the fitness function value for each individual in the population;

[0011] Step 4: Select individuals in the population based on the fitness function value of each individual and a hybrid selection strategy that combines elite preservation and tournament selection. Then, perform crossover and mutation on the selected individuals to obtain the next generation population.

[0012] Step 5: If the maximum number of iterations has been reached, the individual with the largest fitness function value obtained during the entire iteration process will be output as the optimal gas-electricity joint dispatch scheme.

[0013] Otherwise, if the maximum number of iterations has not been reached, continue to execute step three on the population obtained in step four.

[0014] Furthermore, the time dimension parameter includes the hourly index within the scheduling period. Date index within the scheduling period Hourly index within each day of the scheduling cycle Time Period Type Index ;in:

[0015] ; , The total number of days within the scheduling period; , , This represents the total number of hours within the scheduling period. , Representing peak period, Represents the usual time period. Represents the valley period;

[0016] The power station parameters include the power station index. Index of gas types required for power generation Gasification Process Index ;in:

[0017] , The total number of power plants owned by steel companies; , Represents blast furnace gas, Represents coke oven gas, Represents converter gas; , Represents the power generation process, This represents the post-processing stage;

[0018] The gas parameters include hourly output of various types of gas, hourly front-end consumption of various types of gas, hourly amount of various types of gas available for power generation, hourly consumption of various types of gas by each power station, hourly total gas consumption by each power station, maximum gas processing capacity of each power station in one hour, gas storage volume of various types of gas at the end of each hour, maximum storage capacity of various types of gas storage tanks, minimum safe gas storage volume of various types of gas storage tanks, consumption ratio coefficient of various types of gas by each power station, and gas escape rate of each process step.

[0019] The power parameters include the power generation efficiency of each power station, the net power generation of each power station per hour, the charging amount of each power station per hour, the discharging amount of each power station per hour, the upper limit of the power storage capacity of each power station, the total power generation per hour, the power purchase amount per hour, the power demand per hour, and the electricity price for each time period.

[0020] Furthermore, the parameters of the genetic algorithm include population size, number of iterations, crossover rate, mutation rate, elite rate, and tournament size.

[0021] Furthermore, the specific process of step two is as follows:

[0022] Step 21: Read the basic operating parameters of the steel enterprise throughout the entire historical scheduling cycle. The basic operating parameters include the hourly output of various types of gas, the hourly gas consumption of front-end users, the maximum hourly gas processing capacity of each power station, the upper and lower limit capacity of various types of gas storage tanks, and the upper limit of the power station's power energy storage capacity.

[0023] Step 22: Initialize individuals in the population based on constraints on available gas quantity, power plant gas ratio, and power supply and demand balance, combined with differentiated gas allocation strategies during peak and off-peak periods.

[0024] Furthermore, the specific process of step two is as follows:

[0025] Step 221: Design the constraints on the availability of coal gas, the power plant gas ratio, and the power supply and demand balance, respectively.

[0026] Step 222: Based on satisfying all the constraints in Step 221, by adopting a differentiated gas allocation strategy for peak and valley periods, calculate the available power generation of various types of gas and the maximum gas consumption of each power station on an hourly basis.

[0027] Steps 2 and 3: Set the individual code length for:

[0028]

[0029] in, The total number of hours in the scheduling cycle. The total number of power stations;

[0030] when At that time, the first individual generated The element represents the first Purchase electricity per hour ;

[0031] when At that time, the first individual generated The element represents the first The power station in the first Total gas consumption per hour ;

[0032] Based on the calculation results of step 222, generate each individual in the initial population, and then perform dual energy storage constraint verification on each generated individual until an initial population in which each initial individual satisfies the dual energy storage constraint verification is obtained.

[0033] Furthermore, in step 221, the available gas quantity is constrained as follows:

[0034]

[0035] in, Indicates the first Hours Coal-like gas production, Indicates the first Hours Front-end consumption of gas-like materials Indicates the first Hours The amount of coal-like gas that can be used for power generation;

[0036] The power plant gas mix ratio constraint is:

[0037]

[0038] in, Indicates the first The power station in the first Hours Consumption of gas-like materials, Indicates the first Each power station consumes The fixed proportioning coefficient of coal gas-like gas, Indicates the first The power station in the first Total gas consumption per hour;

[0039] And it satisfies:

[0040]

[0041]

[0042] in, Indicates the first The maximum hourly gas processing capacity of each power plant Indicates the first End of hour Gas storage capacity similar to coal gas;

[0043] The constraints on the balance between electricity supply and demand are:

[0044]

[0045] in, For the first Electricity purchased per hour For the first Hourly electricity demand For the first Total hourly power generation Indicates the first The power station in the first Hourly charging capacity;

[0046]

[0047] in, Indicates the first The power station in the first Net power generation per hour Indicates the first The power station in the first Hourly discharge rate;

[0048]

[0049] in, Indicates the first The full-load power generation efficiency of each power station.

[0050] Furthermore, the dual energy storage constraints include gas storage tank capacity constraints and electrical energy storage capacity constraints, wherein:

[0051] The capacity constraint of the gas storage tank is:

[0052]

[0053] in, express End of hour Storage capacity of coal-like gas, express Minimum safe gas storage capacity for gas storage tanks containing coal gas. express The maximum safe gas storage capacity of gas storage tanks for coal-like gases;

[0054]

[0055] in, express End of hour Storage capacity of coal-like gas, express Hours Excess reserves and overflow of coal-like gas;

[0056]

[0057] The energy storage capacity constraint is:

[0058]

[0059] in, Indicates the first The upper limit of the power storage capacity of a single power station For the first The first power station Hourly charging capacity For the first The first power station Hourly discharge rate;

[0060] And it satisfies:

[0061]

[0062]

[0063] .

[0064] Furthermore, the fitness function value is calculated as follows:

[0065]

[0066] in, For the individual fitness function value, This represents the total electricity purchase cost within the dispatch cycle. This represents the equivalent treatment cost of gas venting calculated based on the total gas venting volume.

[0067]

[0068] in, The electricity price for the s-th time period is given, where P represents peak hours, N represents normal hours, and O represents valley hours. Purchase electricity for hour t;

[0069]

[0070] in, Indicates the amount of gas emitted. This indicates the unit price for comprehensive disposal of coal gas venting per unit volume;

[0071]

[0072] in, Let t be the excess storage overflow of Class g gas in hour t. The gas escape rate during power generation. This refers to the rate of gas escape during the post-processing stage.

[0073] Furthermore, the hybrid selection strategy that combines elite retention with tournament selection is as follows: the top 10% of individuals with the highest fitness function values ​​in the population are directly copied and retained, and then the remaining 90% of individuals are determined through tournament selection.

[0074] The crossover process for the selected individuals is as follows: a natural day boundary is randomly selected as the dividing point, and two individuals are randomly selected from all the selected individuals. The two randomly selected individuals are called parent generation 1 and parent generation 2, respectively. The first half of the code of parent generation 1 is combined with the second half of the code of parent generation 2 to generate a new offspring; the second half of the code of parent generation 1 is combined with the first half of the code of parent generation 2 to generate another new offspring.

[0075] The specific process of the mutation operation is as follows: for different time periods, different differential mutation ranges are set for the electricity purchase gene and the gas consumption gene of the power station in the individual.

[0076] Furthermore, different differential variation ranges are set for the electricity purchase gene and the power plant gas consumption gene in individuals, specifically as follows:

[0077] During peak hours:

[0078] Limit the variation range of the electricity purchase gene to The variation range of the gas consumption gene in power plants is limited to... ;

[0079] During valley hours:

[0080] Limit the variation range of the electricity purchase gene to The variation range of the gas consumption gene in power plants is limited to... ;

[0081] in, For the first Hourly electricity demand Indicates the first The power station in the first Hours Consumption of gas-like materials.

[0082] The beneficial effects of this invention are:

[0083] This invention constructs an hourly gas-electricity coupled scheduling model with hourly gas production, front-end gas consumption, electricity demand, time-of-use electricity price, gas storage tanks, and initial power plant status as inputs, and hourly electricity purchase, gas consumption at each power plant, charging and discharging volume, and various gas emissions as outputs. The core optimization objective is to minimize the total electricity purchase cost for steel enterprises, while minimizing gas emissions is used as an auxiliary optimization objective, achieving deep coordination between gas and electricity. Furthermore, the model incorporates gas storage tank capacity constraints and power plant energy storage constraints to improve the gas-electricity dual energy storage constraint system. Finally, the traditional genetic algorithm is improved to enhance convergence efficiency and engineering feasibility. Attached Figure Description

[0084] Figure 1 This is an overall architecture diagram of the gas-electricity joint dispatching system of the present invention;

[0085] Figure 2 This is a convergence curve of the improved genetic algorithm of the present invention;

[0086] Figure 3 This is a graph showing the hourly flow rate and emission rate of blast furnace gas and coke oven gas in an embodiment of the present invention.

[0087] Figure 4 This is a diagram showing the hourly power supply and demand structure within the scheduling period in this embodiment of the invention.

[0088] Figure 5 This is a pie chart showing the power generation ratio of each power station in this embodiment of the invention;

[0089] Figure 6 This is a pie chart showing the proportion of energy storage charging capacity of each power station in this embodiment of the invention. Detailed Implementation

[0090] This invention is based on a constructed hourly gas-electricity coupled scheduling model, which optimizes and dynamically controls the energy flow in the steel production process in real time. The inputs to the hourly gas-electricity coupled scheduling model are real-time operating parameters collected from the steel enterprise's production site, specifically including hourly gas production, front-end gas consumption, electricity demand, time-of-use electricity price, gas storage tanks, and the initial state of the power plant. The outputs include hourly electricity purchases, gas consumption at each power plant, charging and discharging volumes, and various types of gas emissions. The core objective of the model is to minimize the total electricity purchase cost of the steel enterprise while simultaneously minimizing gas emissions. Gas emissions are integrated into the core objective of total electricity purchase cost as a low-weight sub-objective. This core objective allows for prioritizing gas-fired power generation and utilizing energy storage during peak electricity price periods, and increasing gas-fired power generation or switching to energy storage charging during off-peak periods when gas storage tanks are near their capacity. Simultaneously, when gas storage falls below a safety threshold, gas-fired power generation is reduced and externally purchased electricity is increased. Furthermore, the model incorporates gas storage tank capacity constraints and power plant energy storage constraints. The following embodiment proposes an improved genetic algorithm to optimize the solution of the hourly gas-electricity coupled scheduling model, such as... Figure 1 As shown, the method specifically includes the following steps:

[0091] Step 1: Input the time dimension parameters, power plant parameters, gas parameters, power parameters, and genetic algorithm parameters;

[0092] The time dimension parameter includes the hourly index within the scheduling period. Date index within the scheduling period Hourly index within each day of the scheduling cycle Time Period Type Index ;in:

[0093] ; , The total number of days within the scheduling period; , , This represents the total number of hours within the scheduling period. , Representing peak period, Represents the usual time period. Represents the valley period;

[0094] The power station parameters include the power station index. Index of gas types required for power generation Gasification Process Index ;in:

[0095] , The total number of power plants owned by steel companies; , Represents blast furnace gas, Represents coke oven gas, Represents converter gas; , Represents the power generation process, This represents the post-processing stage;

[0096] The gas parameters include hourly output of various types of gas, hourly front-end consumption of various types of gas, hourly amount of various types of gas available for power generation, hourly consumption of various types of gas by each power station, hourly total gas consumption by each power station, maximum gas processing capacity of each power station in one hour, gas storage volume of various types of gas at the end of each hour, maximum storage capacity of various types of gas storage tanks, minimum safe gas storage volume of various types of gas storage tanks, consumption ratio coefficient of various types of gas by each power station, and gas escape rate of each process step.

[0097] The power parameters include the power generation efficiency of each power station, the net power generation of each power station per hour, the charging amount of each power station per hour, the discharging amount of each power station per hour, the upper limit of the power storage capacity of each power station, the total power generation per hour, the power purchase amount per hour, the power demand per hour, and the electricity price for each time period.

[0098] The parameters of the genetic algorithm include population size (preferably 120), number of iterations (preferably 200), crossover rate (preferably 0.85), mutation rate (preferably 0.08), elite rate (preferably 0.1), and tournament size (preferably 3).

[0099] Step 2: Using a real number coding scheme, the hourly electricity purchase and the hourly gas consumption of each power station are mapped into continuous scheduling decision variables. The peak-valley differentiated pre-allocation mechanism and scheduling decision variables are used to generate an initial population. Each individual in the initial population is an hourly scheduling scheme corresponding to the entire scheduling cycle.

[0100] Specifically,

[0101] Step 21: Read the basic operating parameters of the steel enterprise throughout the entire historical scheduling cycle. The basic operating parameters include the hourly output of various types of gas, the hourly gas consumption of front-end users, the maximum hourly gas processing capacity of each power station, the upper and lower limit capacity of various types of gas storage tanks, and the upper limit of the power station's power energy storage capacity.

[0102] Step 22: Initialize individuals in the population based on constraints on available gas quantity, power plant gas ratio, and power supply and demand balance, combined with differentiated gas allocation strategies during peak and off-peak periods;

[0103] The specific process of step two is as follows:

[0104] Step 221: Design the constraints on the availability of coal gas, the power plant gas ratio, and the power supply and demand balance, respectively.

[0105] The gas availability constraint is as follows:

[0106]

[0107] in, Indicates the first Hours Production of coal gas (in ten thousand cubic meters). Indicates the first Hours Front-end consumption of gas-like materials (unit: 10,000 cubic meters). Indicates the first Hours The amount of coal gas that can be used for power generation (in ten thousand cubic meters).

[0108] The total gas consumption of the power plant is automatically broken down into the consumption of various types of gas through a preset ratio. The gas ratio constraint of the power plant is as follows:

[0109]

[0110] in, Indicates the first The power station in the first Hours Consumption of coal gas (in ten thousand cubic meters). Indicates the first Each power station consumes The fixed proportioning coefficient of coal gas-like gas, Indicates the first The power station in the first Total gas consumption per hour (in ten thousand cubic meters).

[0111] And it satisfies:

[0112]

[0113]

[0114] in, Indicates the first Maximum hourly gas processing capacity of each power station (in ten thousand cubic meters). Indicates the first End of hour Gas storage capacity (in ten thousand cubic meters);

[0115] The power supply and demand balance constraint is:

[0116]

[0117] in, For the first Electricity purchased per hour (in MWh). For the first Hourly electricity demand (in MWh). For the first Total hourly power generation (including net gas-fired power generation and energy storage discharge, in MWh) Indicates the first The power station in the first Hourly charging capacity (in MWh);

[0118]

[0119] in, Indicates the first The power station in the first Net power generation per hour (in MWh). Indicates the first The power station in the first Hourly discharge rate;

[0120]

[0121] in, Indicates the first The full-load power generation efficiency of a power station;

[0122] Step 222: Based on satisfying all the constraints in Step 221, by adopting a differentiated gas allocation strategy for peak and valley periods, calculate the available power generation of various types of gas and the maximum gas consumption of each power station on an hourly basis.

[0123] It should be noted that the differentiated gas allocation strategy for peak and off-peak periods is as follows: 30% of the gas storage capacity is reserved. Under the premise of prioritizing the gas storage capacity, an allocation coefficient of 0.9 is adopted during peak periods, that is, 90% of the available gas is allocated to power plants for power generation during peak periods to minimize the purchase of electricity at high prices during peak periods. During off-peak periods, an allocation coefficient of 0.7 is adopted, that is, 70% of the available gas is allocated to power plants for power generation during off-peak periods to reduce the risk of over-storage and release during subsequent peak production periods.

[0124] Steps 2 and 3: Set the individual code length for:

[0125]

[0126] in, The total number of hours in the scheduling cycle. The total number of power stations;

[0127] when At that time, the first individual generated The element represents the first Purchase electricity per hour This is the core variable for adapting to the differentiated peak-valley electricity pricing.

[0128] when At that time, the first individual generated The element represents the first The power station in the first Total gas consumption per hour The core variable for adapting to peak-valley power generation demand;

[0129] Based on the calculation results of step 222, generate each individual in the initial population, and then perform dual energy storage constraint verification on each generated individual until an initial population in which each initial individual satisfies the dual energy storage constraint verification is obtained.

[0130] The dual energy storage constraints include gas storage tank capacity constraints and electrical energy storage capacity constraints, wherein:

[0131] The capacity constraint of the gas storage tank is:

[0132]

[0133] in, express End of hour Storage capacity of coal-like gas, express Minimum safe gas storage capacity (in 10,000 cubic meters) for gas storage tanks containing coal gas. express Maximum safe gas storage capacity of gas storage tanks for coal gas (unit: 10,000 cubic meters).

[0134]

[0135] in, express End of hour Storage capacity of coal gas (unit: 10,000 cubic meters). express Hours Excess reserves and overflow of coal-like gas;

[0136]

[0137] The energy storage capacity constraint is:

[0138]

[0139] in, Indicates the first The upper limit of the power storage capacity of a power station (in MWh). For the first The first power station Hourly charging capacity (in MWh) For the first The first power station Hourly discharge capacity (in MWh);

[0140] And it satisfies:

[0141]

[0142] in, The value is taken as 0.2 times the power generation per hour at full load of the power station;

[0143]

[0144] When the total power generation exceeds the power demand, the power station performs a charging operation;

[0145]

[0146] When the total power generation is insufficient and there is remaining energy storage, the power station will perform a discharge operation.

[0147] Step 3: Construct a dual-constraint fitness function that integrates electricity purchase cost and dual energy storage constraint violation penalty, and calculate the fitness function value for each individual in the population. The calculation method for the fitness function value is as follows:

[0148]

[0149] in, For the individual fitness function value, This represents the total electricity purchase cost within the dispatch cycle (in yuan). This represents the equivalent treatment cost of coal gas emissions (in yuan) calculated based on the total coal gas emissions, with a coefficient. Used to avoid cases where the denominator is zero. This indicates a penalty for violating constraints. A fixed penalty is imposed when an individual violates any gas-side or electricity-side constraint; otherwise, a penalty is imposed when the individual does not violate any gas-side or electricity-side constraint. The value is 0;

[0150]

[0151] in, The electricity price for the s-th time period (in yuan / kWh), where P represents peak period, N represents normal period, and O represents valley period. For the electricity purchased in hour t, the coefficient is... This is used to convert the electricity price unit from yuan / kWh to yuan / MWh, so as to keep it consistent with the unit of electricity purchased;

[0152]

[0153] in, This indicates the amount of gas released (in ten thousand cubic meters). This indicates the unit price for comprehensive disposal of coal gas emissions per unit volume, which includes coal gas energy loss costs, environmental protection costs, and carbon emission conversion costs, and is expressed in yuan per 10,000 cubic meters.

[0154] This invention transforms the amount of coal gas venting into a quantified economic cost through the unit disposal price, thereby achieving the integration of the venting optimization objective and the core objective of electricity purchase cost into the same dimension and uniformly incorporating them into the fitness function evaluation system, eliminating the defects of inconsistent dimensions of multiple objectives and difficulty in labeling weights;

[0155]

[0156] in, The excess storage overflow of Class G gas in hour t (in ten thousand cubic meters). The gas escape rate during power generation. This refers to the rate of gas escape during the post-processing stage.

[0157] Step 4: Select individuals in the population based on the fitness function value of each individual and a hybrid selection strategy that combines elite retention and tournament selection. Then, perform crossover and mutation on the selected individuals to obtain the next generation population. It should be noted that in the process of obtaining the next generation population, it is necessary to perform dual energy storage verification on the newly obtained individuals until all individuals in the next generation population satisfy the dual energy storage constraint.

[0158] Specifically,

[0159] The hybrid selection strategy that combines elite retention with tournament selection is as follows: the top 10% of individuals with the highest fitness function values ​​in the population are directly copied and retained to ensure that the optimal solution is not lost. Then, the remaining 90% of individuals are determined through tournament selection. That is, the scale of the tournament is set to 3. Each time, 3 individuals are randomly selected from the remaining population. The individual with the highest fitness function value among the 3 selected individuals is retained. This tournament selection process is repeated until a sufficient number of individuals are selected.

[0160] The crossover process for the selected individuals is as follows: a natural day boundary is randomly selected as the split point, and the crossover position is the starting index of the code for that day. Then, two individuals are randomly selected from all the individuals. These two randomly selected individuals are called Parent 1 and Parent 2, respectively. The first half of the code of Parent 1 (before the split point) is combined with the second half of the code of Parent 2 (after the split point) to generate a new offspring. The second half of the code of Parent 1 (after the split point) is combined with the first half of the code of Parent 2 (before the split point) to generate another new offspring.

[0161] The specific process of the mutation operation is as follows: For different time periods, different differential mutation ranges are set for the electricity purchase gene and the power plant gas consumption gene in individuals, that is,

[0162] During peak hours:

[0163] Limit the variation range of the electricity purchase gene to Force a reduction in the proportion of high-priced externally purchased electricity; limit the variation range of gas consumption genes in power plants to Allowing for a moderate increase in gas-fired power generation to cover peak electricity demand;

[0164] During valley hours:

[0165] Limit the variation range of the electricity purchase gene to Appropriately increase the substitution of some coal gas power generation with low-priced purchased electricity; limit the variation range of the gas consumption gene of power plants to... Reduce coal-fired power generation to reserve storage capacity.

[0166] Step 5: If the maximum number of iterations has been reached, the individual with the largest fitness function value obtained during the entire iteration process will be output as the optimal gas-electricity joint dispatch scheme, and then the operation of the gas production, gas storage, gas power generation and electricity procurement systems of the steel enterprise will be dispatched.

[0167] Otherwise, if the maximum number of iterations has not been reached, continue to execute step three on the population obtained in step four.

[0168] Example

[0169] This example uses a large steel enterprise with an annual output of 10 million tons as the application object. The enterprise is equipped with 5 gas-fired power plants. The scheduling period is from September 10 to September 16, 2024 (a total of 7 days, or 168 hours). This period avoids the impact of extreme weather and covers the complete production rhythm of the enterprise, which is typical and representative.

[0170] Based on the company's actual operational data, the core model parameters are set as shown in Tables 1, 2, 3, and 4:

[0171] Table 1 Basic Parameter Settings

[0172]

[0173] Table 2 Gas Terminal Parameter Setting Table

[0174]

[0175] Table 3 Power Plant Operating Parameter Setting Table

[0176]

[0177] Table 4. Time-of-use electricity price parameter setting table

[0178]

[0179] To achieve centralized information management and coordinated scheduling of coal gas production, coal gas storage, gas-fired power generation, and electricity procurement, the core optimization objective is to minimize the total electricity purchase cost, while minimizing coal gas emissions is used as an auxiliary optimization objective. A complete constraint system is constructed that includes dual energy storage constraints on both the coal gas and electricity sides.

[0180] The model inputs include hourly gas production, gas demand from front-end users, and electricity demand from enterprises within the scheduling cycle. The data comes from the enterprise's energy management system and production scheduling records, and is processed to ensure continuity after outlier removal and time interpolation.

[0181] The improved genetic algorithm is executed according to the following procedure:

[0182] Population initialization: An initial population is generated using a peak-valley differentiated pre-allocation mechanism. During peak hours, 90% of the available gas is allocated for power generation, while during off-peak hours, only 70% of the available gas is allocated, with storage capacity reserved. All initial individuals are verified under dual energy storage constraints to ensure that the initial solutions meet engineering requirements.

[0183] Fitness calculation: based on the formula Calculate the fitness value for each individual;

[0184] in, This is a penalty for violating the dual energy storage constraint.

[0185] Population evolution:

[0186] (1) Selection

[0187] The top 10% of individuals with the highest fitness in each generation are retained and directly enter the next generation, while a tournament selection (scale 3) is used to select parent individuals;

[0188] (2) Cross

[0189] Randomly select the boundary of a natural day as the dividing point, and perform the day segmentation two-point intersection operation;

[0190] (3) Variation

[0191] Differential mutation operations are performed on the electricity purchase gene and the gas consumption gene of the power plant during peak and valley periods.

[0192] Iteration Termination: When the preset number of iterations (200) is reached, the optimal scheduling scheme corresponding to the individual with the highest fitness function value is output.

[0193] The algorithm convergence process is as follows Figure 2 As shown, when iterated to about 50 generations, the cost of purchased electricity dropped rapidly from the initial 3.6 million yuan to about 2.8 million yuan, and stabilized at 2.7113 million yuan after 170 generations, without premature convergence.

[0194] Analysis of gas utilization efficiency:

[0195] like Figure 3 As shown, the power plant's gas consumption dynamically follows production fluctuations, and the gas storage capacity in the storage tanks is consistently maintained within a safe range of 100,000 to 1,500,000 cubic meters. The total gas release is 413,100 cubic meters, with only a small amount released during individual peak production periods, a reduction of 6.69% compared to traditional manual scheduling.

[0196] Analysis of the effectiveness of power dispatching:

[0197] like Figure 4 As shown, the proportion of purchased electricity during peak hours is significantly reduced. During peak hours (such as the 50th hour), power generation plus energy storage discharge can meet more than 85% of the electricity demand. During off-peak hours, the purchase of low-priced electricity is moderately increased and energy storage is charged, realizing the cost optimization logic of "off-peak charging and peak discharging".

[0198] Power plant load distribution analysis:

[0199] like Figure 5 As shown, the No. 2 South Station (pure coke oven gas combustion), which has the highest power generation efficiency, accounts for 33.7% of the total power generation. Figure 6 As shown, the energy storage charging capacity is evenly distributed among the power stations (24.4% at the North Station and 21.3% at the Central Station), which not only improves energy conversion efficiency but also ensures the stability of system operation.

[0200] In the same operating environment of Windows system and Python 3.10, the method of this invention (IGA) and the basic genetic algorithm (GA) were run 20 times. The average of the optimization objective, the 5% convergence generation, and the average optimization time in the running results were used as comparison indicators for quantitative comparison. At the same time, it was compared with the traditional manual scheduling solution of enterprises. The comparison results are shown in Table 5:

[0201] Table 5 Comparison of Different Scheduling Methods

[0202]

[0203] The comparative results show that the method of the present invention is significantly superior to the existing technology in terms of cost optimization, energy consumption control and algorithm performance, and can meet the engineering requirements of gas-electricity joint dispatching in steel enterprises.

[0204] The above examples of the present invention are merely illustrative of the computational model and process of the present invention, and are not intended to limit the implementation of the present invention. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is impossible to exhaustively list all possible implementations here. Any obvious variations or modifications derived from the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A gas-electricity coordinated optimization scheduling method for steel enterprises based on an improved genetic algorithm, characterized in that, The method specifically includes the following steps: Step 1: Input the time dimension parameters, power plant parameters, gas parameters, power parameters, and genetic algorithm parameters; Step 2: Map the hourly electricity purchase and the hourly gas consumption of each power station into continuous scheduling decision variables. Use the peak-valley differentiated pre-allocation mechanism and scheduling decision variables to generate an initial population. Each individual in the initial population is an hourly scheduling scheme corresponding to the entire scheduling cycle. Step 3: Calculate the fitness function value for each individual in the population; Step 4: Select individuals in the population based on the fitness function value of each individual and a hybrid selection strategy that combines elite preservation and tournament selection. Then, perform crossover and mutation on the selected individuals to obtain the next generation population. Step 5: If the maximum number of iterations has been reached, the individual with the largest fitness function value obtained during the entire iteration process will be output as the optimal gas-electricity joint dispatch scheme. Otherwise, if the maximum number of iterations has not been reached, continue to execute step three on the population obtained in step four.

2. The gas-electricity coordinated optimization scheduling method for steel enterprises based on an improved genetic algorithm according to claim 1, characterized in that, The time dimension parameter includes the hourly index within the scheduling period. Date index within the scheduling period Hourly index within each day of the scheduling cycle Time Period Type Index ;in: ; , The total number of days within the scheduling period; , , This represents the total number of hours within the scheduling period. , Representing peak period, Represents the usual time period. Represents the valley period; The power station parameters include the power station index. Index of gas types required for power generation Gasification Process Index ;in: , The total number of power plants owned by steel companies; , Represents blast furnace gas, Represents coke oven gas, Represents converter gas; , Represents the power generation process, Represents the post-processing stage; The gas parameters include hourly output of various types of gas, hourly front-end consumption of various types of gas, hourly amount of various types of gas available for power generation, hourly consumption of various types of gas by each power station, hourly total gas consumption by each power station, maximum gas processing capacity of each power station in one hour, gas storage volume of various types of gas at the end of each hour, maximum storage capacity of various types of gas storage tanks, minimum safe gas storage volume of various types of gas storage tanks, consumption ratio coefficient of various types of gas by each power station, and gas escape rate of each process step. The power parameters include the power generation efficiency of each power station, the net power generation of each power station per hour, the charging amount of each power station per hour, the discharging amount of each power station per hour, the upper limit of the power storage capacity of each power station, the total power generation per hour, the power purchase amount per hour, the power demand per hour, and the electricity price for each time period.

3. The method for coordinated optimization scheduling of gas and electricity in steel enterprises based on an improved genetic algorithm as described in claim 2, characterized in that, The parameters of the genetic algorithm include population size, number of iterations, crossover rate, mutation rate, elite rate, and tournament size.

4. The method for coordinated optimization scheduling of gas and electricity in steel enterprises based on an improved genetic algorithm as described in claim 1, characterized in that, The specific process of step two is as follows: Step 21: Read the basic operating parameters of the steel enterprise throughout the entire historical scheduling cycle. The basic operating parameters include the hourly output of various types of gas, the hourly gas consumption of front-end users, the maximum hourly gas processing capacity of each power station, the upper and lower limit capacity of various types of gas storage tanks, and the upper limit of the power station's power energy storage capacity. Step 22: Initialize individuals in the population based on constraints on available gas quantity, power plant gas ratio, and power supply and demand balance, combined with differentiated gas allocation strategies during peak and off-peak periods.

5. A gas-electricity coordinated optimization scheduling method for steel enterprises based on an improved genetic algorithm, as described in claim 4, is characterized in that... The specific process of step two is as follows: Step 221: Design the constraints on the availability of coal gas, the power plant gas ratio, and the power supply and demand balance, respectively. Step 222: Based on satisfying all the constraints in Step 221, by adopting a differentiated gas allocation strategy for peak and valley periods, calculate the available power generation of various types of gas and the maximum gas consumption of each power station on an hourly basis. Steps 2 and 3: Set the individual code length for: in, The total number of hours in the scheduling cycle. This represents the total number of power stations; when At that time, the first individual generated The element represents the first Purchase electricity per hour ; when At that time, the first individual generated The element represents the first The power station in the first Total gas consumption per hour ; Based on the calculation results of step 222, generate each individual in the initial population, and then perform dual energy storage constraint verification on each generated individual until an initial population in which each initial individual satisfies the dual energy storage constraint verification is obtained.

6. A gas-electricity coordinated optimization scheduling method for steel enterprises based on an improved genetic algorithm, as described in claim 5, is characterized in that... In step 221, the available gas quantity is constrained as follows: in, Indicates the first Hours Coal-like gas production, Indicates the first Hours Front-end consumption of gas-like materials Indicates the first Hours The amount of coal-like gas that can be used for power generation; The power plant gas mix ratio constraint is: in, Indicates the first The power station in the first Hours Consumption of gas-like substances Indicates the first Each power station consumes The fixed proportioning coefficient of coal gas-like gas, Indicates the first The power station in the first Total gas consumption per hour; And it satisfies: in, Indicates the first The maximum hourly gas processing capacity of each power station Indicates the first End of hour Gas storage capacity similar to coal gas; The constraints on the balance between electricity supply and demand are: in, For the first Electricity purchased per hour For the first Hourly electricity demand For the first Total hourly power generation Indicates the first The power station in the first Hourly charging capacity; in, Indicates the first The power station in the first Net power generation per hour Indicates the first The power station in the first Hourly discharge rate; in, Indicates the first The full-load power generation efficiency of each power station.

7. A gas-electricity coordinated optimization scheduling method for steel enterprises based on an improved genetic algorithm, as described in claim 6, is characterized in that... The dual energy storage constraints include gas storage tank capacity constraints and electrical energy storage capacity constraints, wherein: The capacity constraint of the gas storage tank is: in, express End of the hour Storage capacity of coal-like gas, express Minimum safe gas storage capacity for gas storage tanks containing coal gas. express The maximum safe gas storage capacity of gas storage tanks for coal-like gases; in, express End of the hour Storage capacity of coal-like gas, express Hours Excess reserves and overflow of coal-like gas; The energy storage capacity constraint is: in, Indicates the first The upper limit of the power storage capacity of a single power station For the first The first power station Hourly charging capacity For the first The first power station Hourly discharge rate; And it satisfies: 。 8. A gas-electricity coordinated optimization scheduling method for steel enterprises based on an improved genetic algorithm, as described in claim 7, is characterized in that, The fitness function value is calculated as follows: in, For the individual fitness function value, This represents the total electricity purchase cost within the dispatch cycle. This represents the equivalent treatment cost of gas venting calculated based on the total gas venting volume. in, The electricity price for the s-th time period is given, where P represents peak hours, N represents normal hours, and O represents valley hours. Purchase electricity for hour t; in, Indicates the amount of gas emitted. This indicates the unit price for comprehensive disposal of coal gas venting per unit volume; in, Let t be the excess storage overflow of Class g gas in hour t. The gas escape rate during power generation. This refers to the rate of gas escape during the post-processing stage.

9. A gas-electricity coordinated optimization scheduling method for steel enterprises based on an improved genetic algorithm, as described in claim 1, is characterized in that... The hybrid selection strategy that combines elite retention with tournament selection is as follows: the top 10% of individuals with the highest fitness function values ​​in the population are directly copied and retained, and then the remaining 90% of individuals are determined through tournament selection. The crossover process for the selected individuals is as follows: a natural day boundary is randomly selected as the dividing point, and two individuals are randomly selected from all the selected individuals. The two randomly selected individuals are called parent generation 1 and parent generation 2, respectively. The first half of the code of parent generation 1 is combined with the second half of the code of parent generation 2 to generate a new offspring; the second half of the code of parent generation 1 is combined with the first half of the code of parent generation 2 to generate another new offspring. The specific process of the mutation operation is as follows: for different time periods, different differential mutation ranges are set for the electricity purchase gene and the gas consumption gene of the power station in the individual.

10. A gas-electricity coordinated optimization scheduling method for steel enterprises based on an improved genetic algorithm, as described in claim 9, is characterized in that... The different ranges of differential variation are set for the electricity purchase gene and the power plant gas consumption gene in individuals, specifically: During peak hours: Limit the variation range of the electricity purchase gene to ; Limit the variation range of the gas consumption gene in power plants to ; During valley hours: Limit the variation range of the electricity purchase gene to The variation range of the gas consumption gene in power plants is limited to... ; in, For the first Hourly electricity demand Indicates the first The power station in the first Hours Consumption of gas-like materials.

Images