Coal mine integrated energy system scheduling method based on time segmentation and fusion

Abstract

The invention discloses a coal mine integrated energy system scheduling method based on time segmentation and fusion. The method comprises the following steps: S1, establishing a CMIES optimization model; s2, taking the minimum cost and the minimum waste energy cost as optimization targets of a CMIES optimization model; s3, determining constraint conditions according to an optimization target of the CMIES optimization model and operation requirements of the coal mine integrated energy system; and S4, executing operation on the optimization target by using a constraint multi-objective evolutionary algorithm of time segmentation and fusion. According to the method, an original large-scale search space is reduced into a low-dimensional subspace, so that the optimization difficulty is reduced; designing a differential variation strategy based on dynamic neighborhood search, and balancing global and local search; two single-target auxiliary populations are set, so that the population diversity is improved; the method is high in robustness and can be suitable for coal mine data in different scenes.

Description

Technical Field

[0001] The present invention belongs to the technical field of energy system optimization, and in particular relates to a scheduling method for a coal mine integrated energy system based on time segmentation and fusion. Background Technique

[0002] Integrated Energy System (IES) can effectively integrate various forms of energy, thus playing a crucial role. These systems can improve energy utilization efficiency, reduce energy consumption and carbon emissions through the coordinated optimization of various energies such as electricity, heat, and gas. As a major coal-producing country, the research on CoalMine Integrated Energy System (CMIES) has special significance in this context. Coal resources, as an important part of China's energy structure, will still be the basis of energy supply for a long time. However, a large amount of associated energy will be generated during coal mining, such as mine ventilation air, mine water, geothermal energy, etc. Traditional IES will discard these associated energies, which not only causes waste of resources but also pollutes the natural environment. While CMIES can use various devices, such as heat pumps, to recover these resources, thereby reducing air pollution and minimizing resource waste. Therefore, in order to improve the efficiency and sustainability of coal mine energy utilization, it is urgent to optimize the scheduling of CMIES to accelerate the low-carbon transformation of the coal industry.

[0003] The purpose of CMIES scheduling optimization is to manage the outputs of various devices, optimize energy utilization efficiency, reduce environmental pollution, and meet the energy consumption needs of nearby residents, factories, and others under the premise of meeting various constraints. However, the CMIES scheduling optimization problem has characteristics such as large scale and multiple constraints, making its solution very difficult. Although some existing algorithms for CMIES scheduling can also solve the problem, they all ignore the large-scale characteristic and do not design targeted strategies for this characteristic, thus unable to obtain a more excellent solution. Summary of the Invention

[0004] To solve the above problems, the present invention provides a scheduling method for a coal mine integrated energy system based on time segmentation and fusion, and a constrained multi-objective evolutionary algorithm based on time segmentation and fusion, to more efficiently optimize the scheduling of CMIES.

[0005] The technical solution adopted by the present invention is: a scheduling method for a coal mine integrated energy system based on time segmentation and fusion, including the following steps:

[0006] S1. Establish a CMIES optimization model. The electricity in the CMIES optimization model mainly comes from photovoltaic (PV), wind turbine (WT), power grid (PG), and gas turbine (GT). PV, WT, and GT generate electricity by converting solar energy, wind energy, and natural gas, and are integrated with the PG. The heat energy is mainly generated by the vent oxidation heat pump (VOE), water source heat pump (WSHP), ground source heat pump (GSHP), and air source heat pump (ASHP). VOE, WSHP, GSHP, and ASHP utilize the associated gas in coal mines, mine water, geothermal energy, and air heat to generate heat energy. The absorption chiller (AC) and the electric chiller (EC) provide cooling energy. The electrical energy storage device (ESD) and the thermal energy storage device (TSD) are used to store excess electrical energy and thermal energy.

[0007] S2. Take minimizing cost and minimizing waste energy cost as the optimization objectives of the CMIES optimization model.

[0008] S3. Determine the constraint conditions according to the optimization objectives of the CMIES optimization model and the operation requirements of the coal mine integrated energy system.

[0009] S4. Use the time-slicing and fusion constrained multi-objective evolutionary algorithm to perform operations on the optimization objectives.

[0010] Furthermore, the optimization objectives include minimizing cost and minimizing waste energy cost.

[0011] The minimizing cost includes the purchase cost C buy,t and the equipment maintenance cost C opma,t :

[0012]

[0013] In the formula, T represents the scheduling period, which is adjusted hourly and has a value of 24; α is the cost coefficient, and different devices have different coefficient values; P, H, and Q respectively represent the output electrical energy, heat energy, and cooling energy power of each device; x ESD_out,t and x TSD_out,t respectively represent the discharge and heat release states of the ESD and TSD; x ESD_in,t and x TSD_in,t respectively represent the energy storage states of the ESD and TSD.

[0014] The minimizing waste energy cost includes the waste renewable energy cost and the waste associated energy cost in coal mines:

[0015]

[0016] In the formula, β is the penalty factor for waste energy, P max and H maxRepresent the maximum output electrical power and thermal power of each device respectively.

[0017] Furthermore, the constraint conditions include electrical balance constraint, thermal balance constraint, cold balance constraint, device output constraint, energy conversion constraint, gas turbine ramp constraint, thermal energy storage device constraint, and electrical energy storage device constraint;

[0018] The electrical balance constraint:

[0019]

[0020] In the formula, P load,t represents the electrical load at time t;

[0021] The thermal balance constraint:

[0022]

[0023] In the formula, H load,t represents the thermal load at time t;

[0024] The cold balance constraint:

[0025] Q EC,t +Q AC,t =Q load,t (5)

[0026] In the formula, Q load,t represents the cold load at time t;

[0027] The device output constraint and energy conversion constraint:

[0028]

[0029] In the formula, P max and P min represent the maximum and minimum output electrical powers of each device respectively, and λ represents the energy conversion coefficient of each device;

[0030] The gas turbine ramp constraint:

[0031]

[0032] In the formula, R up and R down represent the maximum and minimum limits of the gas turbine ramp respectively;

[0033] The thermal energy storage device constraint:

[0034]

[0035] In the formula, x ESD_out,t equals 1 and x ESD_in,tWhen it is equal to 0, the device is in a discharging state; x ESD_out,t is equal to 0 and x ESD_in,t When it is equal to 1, the device is in a charging state; x ESD_out,t is equal to 0 and x ESD_in,t When it is equal to 0, the device is in a stationary state; since the device cannot be in the charging and discharging states simultaneously, the two state variables cannot be 1 at the same time. and respectively represent the maximum discharging and charging powers of the device, and respectively represent the maximum and minimum electricity storage limits of the device;

[0036] The constraints of the electricity storage device are as follows:

[0037]

[0038] In the formula, x TSD_out,t is equal to 1 and x TSD_in,t When it is equal to 0, the device is in a heat - releasing state; x TSD_out,t is equal to 0 and x TSD_in,t When it is equal to 1, the device is in a heat - charging state; x TSD_out,t is equal to 0 and x TSD_in,t When it is equal to 0, the device is in a stationary state; since the device cannot be in the heat - charging and heat - releasing states simultaneously, the two state variables cannot be 1 at the same time; and respectively represent the maximum heat - releasing and heat - charging powers of the device, and respectively represent the maximum and minimum heat storage limits of the device.

[0039] Furthermore, S4 includes the following steps:

[0040] S41, Initialize the main population P1 for storing non - dominated solutions, with the population size of NP;

[0041] S42, Initialize the auxiliary population P2 for optimizing the objective f1, with the population size of 1 / 6*NP;

[0042] S43, Initialize the auxiliary population P3 for optimizing the objective f2, with the population size of 1 / 6*NP;

[0043] S44, Record the computing resources used Fe = 4 / 3*NP;

[0044] S45, Set the computing resources fir_Fe = a*Maxfe used in the subspace optimization stage;

[0045] S46. Set the maximum computing resource Maxfe = Maxfe - Fe, and set the currently used computing resource Fe = 0;

[0046] S47. Execute the subspace optimization stage;

[0047] S48. Determine whether the computing resource reaches fir_Fe. If it reaches, execute S49; otherwise, continue to execute S47;

[0048] S49. Execute the original space optimization stage.

[0049] Furthermore, the specific steps of S47 include:

[0050] S471. Set the counter s = 0;

[0051] S472. Set s = s + 1, and determine whether it is greater than Seg. If it is satisfied, output populations P1, P2, and P3; otherwise, jump to S473;

[0052] S473. k = sum(1:s);

[0053] S474. Determine whether s is greater than 1. If it is, randomly initialize half of the individuals in each population; if not, execute S475;

[0054] S475. Execute the offspring generation strategy for the three populations respectively to generate offspring O1, O2, and O3 and evaluate them;

[0055] S476. Set Fe = Fe + 4 / 3 * NP;

[0056] S477. Set O = [O1, O2, O3];

[0057] S478. Based on the two objective function values and the constraint violation degree, use the ∈ - constraint method to select NP individuals from [P1, O] to form a new population P1;

[0058] S479. Based on the f1 objective function value and the constraint violation degree, use the ∈ - constraint method to select NP / 6 individuals from [P2, O] to form a new population P2;

[0059] S480. Based on the f2 objective function value and the constraint violation degree, use the ∈ - constraint method to select NP / 6 individuals from [P3, O] to form a new population P3;

[0060] S481. Determine whether Fe reaches If it reaches, jump to S472; otherwise, jump to S475.

[0061] Furthermore, the specific steps of S49 include:

[0062] S491. Execute the offspring generation strategy for three populations respectively, generate offspring O1, O2, and O3 and evaluate them;

[0063] S492. Set Fe = Fe + 4 / 3 * NP;

[0064] S493. Set O = [O1, O2, O3];

[0065] S494. Based on the two objective function values and the constraint violation degree, use the ε - constraint method to select NP individuals from [P1, O] to form a new population P1;

[0066] S495. Based on the f1 objective function value and the constraint violation degree, use the ε - constraint method to select NP / 6 individuals from [P2, O] to form a new population P2;

[0067] S496. Based on the f2 objective function value and the constraint violation degree, use the ∈ - constraint method to select NP / 6 individuals from [P3, O] to form a new population P3;

[0068] S497. Judge whether the maximum computing resource is reached. If so, output the population P1; otherwise, jump to S491.

[0069] Furthermore, the specific steps of the S475 or the S491 include:

[0070] S4751. Randomly sort the individuals in the population P to obtain rP;

[0071] S4752. Obtain the objective function values of the population rP to get rP obj ;

[0072] S4753. Set rP obj = rP obj - O min , where O min is the minimum objective function value in the population;

[0073] S4754. Calculate the angle between every two individuals based on rP obj to obtain Ang;

[0074] S4755. Set n = 1;

[0075] S4756. Use the following formula to calculate the neighborhood size of the n - th individual in rP

[0076]

[0077] S4757. Randomly select two individuals from the neighbors of the n - th individual: {x r1 , x r2};

[0078] S4758, randomly select a scaling factor F from the parameter pool {0.6, 0.8, 1.0};

[0079] S4759, generate the offspring O using the following formula n :

[0080] O n = x n + F * (x r1 - x r2 )

[0081] S4760, set n = n + 1; if n is greater than NP, output the offspring population O, otherwise, jump to S4756.

[0082] The beneficial effects of the present invention are as follows:

[0083] (1) Reduce the original large-scale search space to a low-dimensional subspace, reducing the optimization difficulty;

[0084] (2) Design a differential mutation strategy based on dynamic neighborhood search to balance global and local search;

[0085] (3) Set two single-objective auxiliary populations to improve population diversity;

[0086] (4) The present invention has strong robustness and can be applied to coal mine data under different scenarios. Description of the Drawings

[0087] Figure 1 is the CMIES optimization model of the present invention;

[0088] Figure 2 is the schematic diagram of the optimization method in the present invention. Detailed Embodiment

[0089] The present invention will be further described below with reference to the drawings.

[0090] The present invention is a scheduling method for a coal mine integrated energy system based on time segmentation and fusion, including the following steps:

[0091] S1, establish a CMIES optimization model; as Figure 1As shown in the figure, the power in the CMIES optimization model mainly comes from photovoltaic PV, wind turbine WT, power grid PG, and gas turbine GT. Photovoltaic PV, wind turbine WT, and gas turbine GT generate electricity by converting solar energy, wind energy, and natural gas, and are integrated with the power grid PG; the heat energy is mainly generated by the waste air oxidation heat pump VOE, water source heat pump WSHP, ground source heat pump GSHP, and air source heat pump ASHP. The waste air oxidation heat pump VOE, water source heat pump WSHP, ground source heat pump GSHP, and air source heat pump ASHP utilize the associated gas, mine water, geothermal energy, and air heat in coal mines to generate heat energy; the absorption chiller AC and the electric chiller EC provide cooling energy; the electricity storage device ESD and the heat storage device TSD are used to store excess electric energy and heat energy to cope with load fluctuations.

[0092] S2. Take minimizing cost and minimizing waste energy cost as the optimization objectives of the CMIES optimization model;

[0093] The minimized cost includes the purchase cost C buy,t and the equipment maintenance cost C opma,t :

[0094]

[0095] In the formula, T represents the scheduling period, which is adjusted once per hour and has a value of 24. α is the cost coefficient, and different devices have different coefficient values. P, H, and Q respectively represent the output electric energy, heat energy, and cooling energy power of each device. x ESD_out,t and x TSD_out,t respectively represent the discharge and heat release states of ESD and TSD. x ESD_in,t and x TSD_in,t respectively represent the energy storage states of the two devices;

[0096] The minimized waste energy cost includes the waste renewable energy cost and the waste associated energy cost in coal mines:

[0097]

[0098] In the formula, β is the penalty factor for waste energy. P max and H max respectively represent the maximum output electric energy power and heat energy power of each device.

[0099] S3. Determine the constraint conditions according to the optimization objectives of the CMIES optimization model and the operation requirements of the coal mine integrated energy system, including power balance constraint, heat balance constraint, cold balance constraint, device output constraint, energy conversion constraint, gas turbine ramp constraint, heat storage device constraint, and electricity storage device constraint;

[0100] Power balance constraint:

[0101]

[0102] Wherein, P load,t represents the electrical load at time t;

[0103] Thermal balance constraint:

[0104]

[0105] Wherein, H load,t represents the thermal load at time t;

[0106] The cold balance constraint:

[0107] Q EC,t +Q AC,t =Q load,t (5)

[0108] Wherein, Q load,t represents the cold load at time t;

[0109] The device output constraint and energy conversion constraint:

[0110]

[0111] Wherein, P max and P min respectively represent the maximum and minimum output electric power of each device, and λ represents the energy conversion coefficient of each device;

[0112] The gas turbine ramp constraint:

[0113]

[0114] Wherein, R up and R down respectively represent the maximum and minimum limits of the gas turbine ramp;

[0115] The heat storage device constraint:

[0116]

[0117] Wherein, x ESD_out,t equals 1 and x ESD_in,t equals 0, the device is in the discharging state; x ESD_out,t equals 0 and x ESD_in,t equals 1, the device is in the charging state; x ESD_out,t equals 0 and x ESD_in,t equals 0, the device is in the static state; since the device cannot be in the charging and discharging states simultaneously, the two state variables cannot be 1 simultaneously. and respectively represent the maximum discharging and charging powers of the device, and represent the maximum and minimum electricity storage limits of the device, respectively;

[0118] The electricity storage device constraints:

[0119]

[0120] In the formula, x TSD_out,t equals 1 and x TSD_in,t equals 0, the device is in the exothermic state; when x TSD_out,t equals 0 and x TSD_in,t equals 1, the device is in the heat charging state; when x TSD_out,t equals 0 and x TSD_in,t equals 0, the device is in the stationary state; since the device cannot be in the heat charging and exothermic states simultaneously, the two state variables cannot be 1 at the same time. and represent the maximum exothermic and heat charging powers of the device, respectively, and represent the maximum and minimum heat storage limits of the device, respectively.

[0121] S4. Use the constrained multi-objective evolutionary algorithm with time segmentation and fusion to perform operations on the optimization objectives, including the following steps:

[0122] S41. Initialize the main population P1 for saving non-dominated solutions, with the population size of NP;

[0123] S42. Initialize the auxiliary population P2 for optimizing the objective f1, with the population size of 1 / 6*NP;

[0124] S43. Initialize the auxiliary population P3 for optimizing the objective f2, with the population size of 1 / 6*NP;

[0125] S44. Record the computing resources used Fe = 4 / 3*NP;

[0126] S45. Set the computing resources fir_Fe = a*Maxfe used in the subspace optimization stage;

[0127] S46. Set the maximum computing resources Maxfe = Maxfe - Fe, and the currently used computing resources Fe = 0;

[0128] S47. Execute the subspace optimization stage; the specific steps include:

[0129] S471. Set the counter s = 0;

[0130] S472. Set s = s + 1, and judge whether it is greater than Seg; if satisfied, output the populations P1, P2, and P3, otherwise, jump to S473;

[0131] S473, k = sum(1:s);

[0132] S474, Determine whether s is greater than 1; if so, randomly initialize half of the individuals in each population; if not, execute S475;

[0133] S475, Execute the offspring generation strategy for each of the three populations to generate offspring O1, O2, and O3 and evaluate them; the specific steps are as follows:

[0134] S4751, Randomly sort the individuals in population P to obtain rP;

[0135] S4752, Obtain the objective function values of population rP to get rP obj ;

[0136] S4753, Set rP obj = rP obj - O min where O min is the minimum objective function value in the population;

[0137] S4754, Calculate the angle between every two individuals based on rP obj to obtain Ang;

[0138] S4755, Set n = 1;

[0139] S4756, Calculate the neighborhood size of the nth individual in rP using the following formula

[0140]

[0141] S4757, Randomly select two individuals from the neighbors of the nth individual: {x r1 , x r2};

[0142] S4758, Randomly select a scaling factor F from the parameter pool {0.6, 0.8, 1.0};

[0143] S4759, Generate offspring O n using the following formula:

[0144] O n = x n + F * (x r1 - x r2 )

[0145] S4760, Set n = n + 1; if n is greater than NP, output the offspring population O, otherwise, jump to S4756;

[0146] S476, set Fe = Fe + 4 / 3 * NP;

[0147] S477, set O = [O1, O2, O3];

[0148] S478, based on two objective function values and constraint violation degree, use the ∈ - constraint method to select NP individuals from [P1, O] to form a new population P1;

[0149] S479, based on the f1 objective function value and constraint violation degree, use the ∈ - constraint method to select NP / 6 individuals from [P2, O] to form a new population P2;

[0150] S480, based on the f2 objective function value and constraint violation degree, use the ∈ - constraint method to select NP / 6 individuals from [P3, O] to form a new population P3;

[0151] S481, judge whether Fe reaches If it reaches, jump to S472; otherwise, jump to S475;

[0152] S48, judge whether the computing resources reach fir_Fe. If they reach, execute S49; otherwise, continue to execute S47;

[0153] S49, execute the original space optimization stage. The specific steps include:

[0154] S491, execute the offspring generation strategy for the three populations respectively, generate offspring O1, O2 and O3 and evaluate them. The specific steps include:

[0155] S4911, randomly sort the individuals in population P to get rP;

[0156] S4912, obtain the objective function values of population rP to get rP obj ;

[0157] S4913, set rP obj = rP obj - O min , where O min is the minimum objective function value in the population;

[0158] S4914, calculate the angle between every two individuals based on rP obj to get Ang;

[0159] S4915, set n = 1;

[0160] S4916, use the following formula to calculate the neighborhood size of the nth individual in rP

[0161]

[0162] S4917, randomly select two individuals from the neighbors of the n-th individual: {x r1 , x r2};

[0163] S4918, randomly select a scaling factor F from the parameter pool {0.6, 0.8, 1.0};

[0164] S4919, generate the offspring O using the following formula n :

[0165] O n = x n + F * (x r1 - x r2 )

[0166] S4920, set n = n + 1; if n is greater than NP, output the offspring population O, otherwise, jump to S4916;

[0167] S492, set Fe = Fe + 4 / 3 * NP;

[0168] S493, set O = [O1, O2, O3];

[0169] S494, based on the two objective function values and the constraint violation degree, use the ε-constraint method to select NP individuals from [P1, O] to form a new population P1;

[0170] S495, based on the f1 objective function value and the constraint violation degree, use the ε-constraint method to select NP / 6 individuals from [P2, O] to form a new population P2;

[0171] S496, based on the f2 objective function value and the constraint violation degree, use the ∈-constraint method to select NP / 6 individuals from [P3, O] to form a new population P3;

[0172] S497, determine whether the maximum computing resources have been reached. If so, output the population P1. Otherwise, jump to S491.

[0173] To verify the effectiveness of the present invention, the present invention is used to optimize the data of a real coal mine in Shanxi. The relevant data can be found in the literature "C.Y. Dai, X.Y. Sun, H.J. Hu, W. Song, Y. Zhang, D.W. Gong. Multiform Differential Evolution With Elite-Guided Knowledge Transfer for Coal Mine Integrated Energy Systems Constrained Dispatch[J]. IEEE Transactions on Evolutionary Computation, doi:10.1109 / TEVC.2024.3496852.". Among them, the specific parameters of the present invention are as follows: the size NP of the main population P1 is set to 300, the sizes of the two auxiliary populations P2 and P3 are NP / 6, that is, 50, and the number Seg of the decomposed sub-problems is 24; the minimum value of the dynamic neighborhood scale is set to 10; the total number of evaluations is 216000, and the proportion a of the maximum number of evaluations occupied in the subspace optimization stage is 0.1.

[0174] The optimization method of the present invention is compared with existing centralized methods, namely BiCo, CCMO, CMOCSO, LCMOEA, SFEA, and MDE_EKT_NDM, respectively. All algorithms are run 20 times to obtain statistical results. Among them, BiCo is the algorithm disclosed in the literature "Z.Z. Liu, B.C. Wang, K.Tang. Handling Constrained Multiobjective Optimization Problems via Bidirectional Coevolution[J]. IEEE Transactions on Cybernetics, 2022, 52(10): 10163-10176."; CCMO is the algorithm disclosed in the literature "Y. Tian, T. Zhang, J, H, Xiao, X.Y. Zhang, Y.C. Jin. A Coevolutionary Framework for Constrained Multiobjective Optimization Problems[J]. IEEE Transactions on Evolutionary Computation, 2021, 25(1): 102-116."; CMOCSO is the algorithm disclosed in the literature "F. Ming, W.Y. Gong, D.C. Li, L. Wang, L.Gao. A Competitive and Cooperative Swarm Optimizer for Constrained Multiobjective Optimization Problems[J]. IEEE Transactions on Evolutionary Computation, 2023, 27(5): 1313-1326."; LCMOEA is the algorithm disclosed in the literature "S.B. Liu, Z.Y. Wang, Q.Z. Lin, J.Q. Li, K.C. Tan. Learning-Aided Evolutionary Search and Selection for Scaling-up Constrained Multiobjective Optimization[J]. IEEE Transactions on Evolutionary Computation, doi:10.1109 / TEVC.2024.3380366."; SFEA is the algorithm disclosed in the literature "J.L. Zhou, Y.G. Zhang, F. Yu, X. Yang, P.N. Suganthan.The algorithm disclosed in "A Staged Fuzzy Evolutionary Algorithm for Constrained Large-Scale Multiobjective Optimization [J]. Applied SoftComputing, 2024, 167, 112297."; MDE_EKT_NDM is the algorithm disclosed in the literature "C.Y. Dai, X.Y. Sun, H.J. Hu, W.Song, Y.Zhang, D.W. Gong. Multiform Differential Evolution With Elite-Guided Knowledge Transfer for Coal Mine Integrated Energy Systems Constrained Dispatch [J]. IEEE Transactions on Evolutionary Computation, doi: 10.1109 / TEVC.2024.3496852.".

[0175] The Inverted Generational Distance (IGD) values and Hypervolume (HV) obtained by the algorithms are shown in Table 1 and Table 2 respectively. Table 1 Comparison results of the IGD index of the present invention and six other advanced evolutionary algorithms for the optimization problem of coal mine integrated energy systems

[0176]

[0177]

[0178] Table 2 Comparison results of the HV index of the present invention and six other advanced evolutionary algorithms for the optimization problem of coal mine integrated energy systems

[0179] Algorithm Worst value Best value Average value Variance BiCo 0.550134 0.82167 0.683823 0.079284 CCMO 0.313432 0.719805 0.508001 0.112105 CMOCSO 0.610717 0.726391 0.666551 0.033782 LCMOEA 0.546367 0.753759 0.663458 0.07052 SFEA 0.517516 0.726595 0.602128 0.063112 MDE_EKT_NDM 0.92397 0.976095 0.952067 0.018337 The method of the present invention 0.999664 1.013603 1.006149 0.003831

[0180] Among them, the smaller the IGD value or the larger the HV, the better the algorithm performance. Analyzing the data in Table 1 and Table 2, it can be seen that the method of the present invention has achieved better results than the six comparison methods in terms of the four indicators of the worst value, best value, average value, and variance. This is attributed to the proposed problem decomposition strategy, differential evolution mutation strategy based on dynamic neighborhood, and single-objective auxiliary population strategy, which effectively improve the diversity and convergence of the population, enabling the algorithm to obtain better results than other algorithms. In actual production operations, factories can fully integrate these solutions according to their own situations, so as to plan a satisfactory energy dispatch and allocation plan.

[0181] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Any other modifications or equivalent replacements made by those of ordinary skill in the art to the technical solutions of the present invention should be covered within the scope of the claims of the present invention as long as they do not depart from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for dispatching a coal mine integrated energy system based on time segmentation and fusion, characterized in that: The following steps are involved: S1, establish a CMIES optimization model; the electricity in the CMIES optimization model mainly comes from photovoltaic PV, wind turbine WT, power grid PG and gas turbine GT, photovoltaic PV, wind turbine WT and gas turbine GT generate electricity by converting solar energy, wind energy and natural gas, and integrate with power grid PG; thermal energy is mainly generated by exhaust air oxidation heat pump VOE, water source heat pump WSHP, ground source heat pump GSHP and gas source heat pump ASHP, exhaust air oxidation heat pump VOE, water source heat pump WSHP, ground source heat pump GSHP and gas source heat pump ASHP use coal mine associated gas, mine water, geothermal energy and air heat to generate thermal energy; absorption refrigerator AC and electric refrigerator EC provide cold energy; electric storage device ESD and heat storage device TSD are used to store excess electric energy and thermal energy; S2, minimizing the cost and minimizing the waste energy cost are taken as the optimization objectives of the CMIES optimization model; S3, determine the constraint conditions according to the optimization objectives of the CMIES optimization model and the operation requirements of the coal mine integrated energy system; S4, uses the constrained multi-objective evolutionary algorithm with time segmentation and fusion to perform operations on the optimization objectives.

2. The method for dispatching a coal mine integrated energy system based on time segmentation and fusion according to claim 1, characterized in that: The optimization objectives include minimizing cost and minimizing waste energy cost; The minimization cost includes the purchase cost C buy,t and equipment maintenance cost C opma,t : Where T represents the scheduling period; α is the cost coefficient; P, H and Q represent the output power of each device, heat energy and cooling energy respectively; x ESD_out,t and x TSD_out,t Respectively represent the discharge and heat release states of ESD and TSD; x ESD_in,t and x TSD_in,t Represent the energy storage states of ESD and TSD respectively; The minimization of the waste energy cost includes the waste renewable energy cost and the waste coal mine associated energy cost: Where β is the penalty factor for wasted energy; P max and H max Represents the maximum output electrical power and thermal power of each device respectively.

3. The method for dispatching a coal mine integrated energy system based on time segmentation and fusion according to claim 1, characterized in that: The constraints include electrical balance constraints, thermal balance constraints, cold balance constraints, device output constraints and energy conversion constraints, gas turbine ramp constraints, thermal storage device constraints and electrical storage device constraints; The electrical balance constraint: Where P load,t represents the electrical load at time t; The thermal balance constraint: In the formula, H load,t represents the heat load at time t; The cold balance constraint: Q EC,t +Q AC,t =Q load,t (5) In the formula, Q load,t represents the cooling load at time t; The device output constraints and energy conversion constraints are: Where P max and P min Respectively represent the maximum and minimum output electrical power of each device; λ represents the energy conversion factor of each device; The gas turbine ramping constraint: In the formula, R up and R down They represent the maximum and minimum limits of the gas turbine ramping, respectively; The heat storage device constraints: In the formula, x ESD_out,t is equal to 1 and x ESD_in,t When x is equal to 0, the device is in the discharge state; ESD_out,t is equal to 0 and x ESD_in,t When it is equal to 1, the device is in charging state; x ESD_out,t is equal to 0 and x ESD_in,t When x is equal to 0, the device is at rest; ESD_out,t and x ESD_in,t Cannot be 1 at the same time; and Respectively represent the maximum discharge and charge power of the device; and Respectively represent the maximum and minimum power storage limits of the device; The power storage device constraints: In the formula, x TSD_out,t is equal to 1 and x TSD_in,t When x is equal to 0, the device is in the exothermic state; TSD_out,t is equal to 0 and x TSD_in,t When x is equal to 1, the device is in the hot state; TSD_out,t is equal to 0 and x TSD_in,t When x is equal to 0, the device is at rest; TSD_out,t and x TSD_in,t Cannot be 1 at the same time; and Respectively represent the maximum heat release and heat charging power of the device; and Represent the maximum and minimum heat storage limits of the device respectively.

4. The method for dispatching a coal mine integrated energy system based on time segmentation and fusion according to claim 1, characterized in that: The S4 comprises the following steps: S41, initialize the main population P1 for storing non-dominated solutions, the population size is NP; S42, initialize the auxiliary population P2 for optimizing the target f1, the population size is 1 / 6*NP; S43, initialize the auxiliary population P3 for optimizing the target f2, the population size is 1 / 6*NP; S44, record the used computing resources Fe=4 / 3*NP; S45, setting the computing resources used in the subspace optimization phase fir_Fe=a*Maxfe; S46, setting the maximum computing resource Maxfe=Maxfe-Fe, and the currently used computing resource Fe=0; S47, executing the subspace optimization phase; S48, determine whether the computing resources have reached fir_Fe, if so, execute S49, otherwise, continue to execute S47; S49, executing the original space optimization phase.

5. A method for dispatching a coal mine integrated energy system based on time segmentation and fusion according to claim 4, characterized in that: The specific steps of S47 include: S471, set counter s=0; S472, set s=s+1, and determine whether it is greater than Seg; if satisfied, output populations P1, P2 and P3, otherwise, jump to S473; S473, k = sum(1:s); S474, determine whether s is greater than 1; if so, randomly initialize half of the individuals in each population; if not, execute S475; S475, executing the offspring generation strategy for the three populations respectively, generating offspring O1, O2 and O3 and evaluating them; S476, set Fe=Fe+4 / 3*NP; S477, set O = [O1, O2, O3]; S478, based on the two objective function values ​​and constraint violation degree, use the ∈ constraint method to select NP individuals in [P1,O] to form a new population P1; S479, based on the f1 objective function value and constraint violation degree, use the ∈ constraint method to select NP / 6 individuals from [P2,O] to form a new population P2; S480, based on the f2 objective function value and constraint violation degree, use the ∈ constraint method to select NP / 6 individuals from [P3,O] to form a new population P3; S481, determine whether Fe reaches If reached, jump to S472; otherwise, jump to S475.

6. A method for dispatching a coal mine integrated energy system based on time segmentation and fusion according to claim 4, characterized in that The specific steps of S49 include: S491, execute the offspring generation strategy for the three populations respectively, generate offspring O1, O2 and O3 and evaluate them; S492, set Fe=Fe+4 / 3*NP; S493, set O = [O1, O2, O3]; S494, based on the two objective function values ​​and constraint violation degree, use the ∈ constraint method to select NP individuals in [P1,O] to form a new population P1; S495, based on the f1 objective function value and constraint violation degree, use the ∈ constraint method to select NP / 6 individuals from [P2,O] to form a new population P2; S496, based on the f2 objective function value and constraint violation degree, use the ∈ constraint method to select NP / 6 individuals from [P3,O] to form a new population P3; S497, determine whether the maximum computing resources are reached, if reached, output population P1, otherwise, jump to S491.

7. A method for dispatching a coal mine integrated energy system based on time segmentation and fusion according to any one of claims 5 or 6, characterized in that The specific steps of S475 or S491 include: S4751, randomly sort the individuals in population P to obtain rP; S4752, obtain the objective function value of the population rP, and obtain rP obj ; S4753, set rP obj =rP obj -O min , where O min is the minimum objective function value in the population; S4754, according to rP obj Calculate the angle between every two individuals and get Ang; S4755, set n=1; S4756, use the following formula to calculate the neighborhood size of the nth individual in rP S4757, randomly select two individuals from the neighbors of the nth individual: {x r1 ,x r2 }; S4758, randomly select a scaling factor F from the parameter pool {0.6, 0.8, 1.0}; S4759, use the following formula to generate progeny O n : O n =x n +F*(x r1 -x r2 ) S4760, set n=n+1; if n is greater than NP, output the offspring population O, otherwise, jump to S4756.

Images