A multi-objective optimization planning method and system for deep peak shaving transformation of thermal power units

Abstract

The present disclosure belongs to the technical field of power system optimization, and specifically relates to a multi-objective optimization planning method and system for deep peak-shaving transformation of thermal power units, comprising: obtaining system parameters of the thermal power units; constructing a multi-objective optimization planning model for deep peak-shaving transformation of the thermal power units based on the obtained system parameters of the thermal power units; solving the constructed multi-objective optimization planning model for deep peak-shaving transformation of the thermal power units to obtain an optimization plan for deep peak-shaving transformation of the thermal power units; wherein the constructed multi-objective optimization planning model for deep peak-shaving transformation of the thermal power units includes an upper-level model and a lower-level model; the upper-level model takes minimizing the total cost of deep peak-shaving and maximizing the deep peak-shaving capacity as the upper-level objective function, and takes the minimum load of the thermal power unit after transformation and optimization as the decision variable; the lower-level model takes minimizing the total operating cost of the unit as the lower-level objective function, and takes the unit output power and the node voltage phase angle as the decision variables.

Description

Technical Field

[0001] The present disclosure belongs to the technical field of power system optimization, and in particular relates to a multi-objective optimization planning method and system for deep peak-shaving transformation of thermal power units. Background Art

[0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art.

[0003] Renewable energy generation is characterized by intermittency and volatility. This is impacted by multiple factors, including a mismatch between load power demand and renewable energy generation, poor transmission capacity, and limited grid regulation. This creates significant peak-shaving pressure on the power system, leading to severe wind and solar curtailment. To improve the absorption capacity of renewable energy and alleviate the grid's peak-shaving pressure during off-peak periods, the peak-shaving capabilities of thermal power units need to be further explored. This means that thermal power units must undergo a transformation and upgrade toward "deep peak-shaving," gradually shifting from a primary power source to an auxiliary service source to provide grid power support and assurance.

[0004] Deep peak-shaving retrofits for thermal power units require a trade-off between affordability and flexibility. Total retrofit costs vary significantly between units, with varying unit characteristics, retrofit objectives, and fuel consumption factors all contributing significantly to the investment required. In terms of total operating costs, the generation costs of thermal power units operating under deep peak-shaving conditions are highly variable. In addition to the basic unit coal consumption, these costs also include additional unit oil consumption and lifespan losses caused by wide load factor fluctuations. Furthermore, the increased environmental impact of retrofit operations translates to higher environmental compensation costs. Furthermore, the potential for deep peak-shaving revenue from retrofitted units decreases, leading to a reasonable peak-shaving compensation fee as a proxy for the benefits of deep peak-shaving. Regarding flexibility, rational site planning and optimal retrofit solutions are crucial to ensure that the post-retrofit deep peak-shaving capacity of the units meets system requirements while maintaining the maximum possible capacity margin to meet future peak-shaving demands. Therefore, in-depth research on deep peak-shaving retrofit solutions for thermal power units is necessary. Currently, most studies addressing this issue rely on simplified models, ignoring the impact of key factors on the effectiveness of thermal power unit retrofits. Summary of the Invention

[0005] In order to solve the above problems, the present disclosure proposes a multi-objective optimization planning method and system for deep peak-shaving transformation of thermal power units. The present disclosure constructs a multi-objective optimization model that takes into account both economy and flexibility. In terms of economy, it considers both transformation cost and operating cost. In terms of flexibility, it strives to maximize the deep peak-shaving capacity of the thermal power units after transformation, achieves a dynamic balance between economy and flexibility in the deep peak-shaving transformation plan of the thermal power units, maintains the optimal deep peak-shaving capacity under the premise of economy, alleviates the peak-shaving pressure of the power grid and absorbs as much new energy as possible.

[0006] According to some embodiments, the first solution of the present disclosure provides a multi-objective optimization planning method for deep peak-shaving transformation of thermal power units, which adopts the following technical solutions:

[0007] A multi-objective optimization planning method for deep peak-shaving transformation of thermal power units includes the following steps:

[0008] Obtain system parameters of thermal power units;

[0009] Based on the obtained system parameters of thermal power units, a multi-objective optimization planning model for deep peak-shaving transformation of thermal power units is constructed;

[0010] Solve the constructed multi-objective optimization planning model for deep peak-shaving transformation of thermal power units and obtain the optimization plan for deep peak-shaving transformation of thermal power units;

[0011] Among them, the constructed multi-objective optimization planning model for deep peak-shaving transformation of thermal power units includes an upper-level model and a lower-level model; the upper-level model takes the minimum total cost of deep peak-shaving and the maximum deep peak-shaving capacity as the upper-level objective function, and the minimum load of the thermal power unit after transformation and optimization as the decision variable; the lower-level model takes the minimum total operating cost of the unit as the lower-level objective function, and the unit output power and node voltage phase angle as the decision variables.

[0012] As a further technical limitation, the obtained thermal power unit system parameters at least include the life of the thermal power unit, the minimum load of each unit, the operating cost of the unit and the coal consumption of the thermal power unit.

[0013] As a further technical limitation, the total cost of deep peak regulation includes the total unit transformation cost and the total unit operating cost; wherein the total unit transformation cost f1 is:

[0014]

[0015] Where N is the total number of thermal power units in the power system to be transformed; S build is the unit deep peak load capacity transformation cost; r is the annual interest rate; y i is the life of the i-th unit; P i，mina is the minimum load of the i-th unit before transformation; P i，minis the minimum load of the i-th unit after transformation, when P i，min Equal to P i，mina When , it means that the i-th unit has not been modified;

[0016] The deep peak-shaving capacity is the difference between the sum of the minimum loads of all thermal power units in the system before the transformation and the sum of the minimum loads of all thermal power units in the system after the transformation, that is, the deep peak-shaving capacity F2 is:

[0017]

[0018] The constraints of the upper objective function include the minimum load constraint of the unit after transformation and the deep peak-shaving capacity constraint of the system after transformation;

[0019] The minimum load constraint of the unit after the transformation is:

[0020] P i，minc ≤P i，min ≤P i，mina

[0021] Among them, P i，minc is the minimum load that the i-th unit can achieve after the transformation;

[0022] The deep peak-shaving capacity constraint of the system after the transformation is:

[0023]

[0024] Among them, P bulild It represents the minimum demand for deep peak-shaving capacity of the power system to be transformed, provided by the power system dispatching department.

[0025] As a further technical limitation, the total operating cost of the unit includes the unit coal consumption cost, oil input fuel cost, life loss cost, environmental compensation cost and peak load compensation cost;

[0026] The constraints of the lower objective function include ramp constraints, unit output constraints, node voltage phase angle constraints, power balance constraints and branch power flow constraints;

[0027] The climbing constraint is:

[0028]

[0029] Among them, P i，t-1 A represents the output power of the i-th unit in the t-1 period; i represents the maximum ramp rate of the i-th unit;

[0030] The unit output constraint is:

[0031] P i，min ≤P i，t≤P i，N

[0032] Among them, P i，min represents the minimum load of the i-th unit after transformation, P i，t represents the output power of the i-th unit in the t-th period, P i，N represents the rated load of the i-th unit;

[0033] The node voltage phase angle constraint is:

[0034] θ j，min ≤θ j，t ≤θ j，max

[0035] Among them, θ j，min ,θ j，max They represent the minimum and maximum values of the node voltage phase angle of the jth node, θ j，t represents the voltage phase angle of the jth node in the tth period;

[0036] The power balance constraint is:

[0037]

[0038] Among them, P j，t represents the output power of the jth node in the tth period; L j represents the load of the jth node; n represents the total number of nodes connected to the jth node; θ k，t and X k They represent the voltage phase angle of node k connected to the jth node and the line impedance between nodes respectively;

[0039] The branch power flow constraint is:

[0040]

[0041] in, and They represent the maximum power flow limit and impedance value of the branch between the j1th node and the j2th node respectively; They represent the voltage phase angles of nodes j1 and j2 in the tth period respectively.

[0042] As a further technical limitation, in the process of solving the constructed multi-objective optimization planning model for deep peak-shaving transformation of thermal power units, a multi-objective particle swarm algorithm based on a periodic elite competition strategy is used to solve the upper-level model, and the lower-level model is solved based on a nonlinear programming solution method to obtain the Pareto optimal solution set and the deep peak-shaving transformation plan for the thermal power units.

[0043] Furthermore, according to the minimum load of the modified unit obtained by solving the upper model, the lower model is solved based on the nonlinear programming solution method to optimize the total operating cost of the unit, and the obtained total operating cost of the unit is fed back to the upper model.

[0044] Furthermore, according to the total operating cost of the unit obtained by solving the lower model, the upper model is solved using a multi-objective particle swarm algorithm based on a periodic elite competition strategy, the Pareto optimal solution set is solved, and the deep peak-shaving transformation plan of the thermal power unit is obtained.

[0045] According to some embodiments, a second solution of the present disclosure provides a multi-objective optimization planning system for deep peak-shaving transformation of thermal power units, which adopts the following technical solutions:

[0046] A multi-objective optimization planning system for deep peak-shaving transformation of thermal power units, comprising:

[0047] an acquisition module configured to acquire system parameters of a thermal power unit;

[0048] A modeling module is configured to construct a multi-objective optimization planning model for deep peak-shaving transformation of thermal power units based on the acquired thermal power unit system parameters;

[0049] an optimization module configured to solve the constructed multi-objective optimization planning model for deep peak-shaving transformation of thermal power units to obtain an optimization plan for deep peak-shaving transformation of thermal power units;

[0050] Among them, the constructed multi-objective optimization planning model for deep peak-shaving transformation of thermal power units includes an upper-level model and a lower-level model; the upper-level model takes the minimum total cost of deep peak-shaving and the maximum deep peak-shaving capacity as the upper-level objective function, and the minimum load of the thermal power unit after transformation and optimization as the decision variable; the lower-level model takes the minimum total operating cost of the unit as the lower-level objective function, and the unit output power and node voltage phase angle as the decision variables.

[0051] According to some embodiments, a third solution of the present disclosure provides a computer-readable storage medium, which adopts the following technical solution:

[0052] A computer-readable storage medium stores a program thereon, which, when executed by a processor, implements the steps of the multi-objective optimization planning method for deep peak-shaving transformation of thermal power units as described in the first aspect of the present disclosure.

[0053] According to some embodiments, a fourth solution of the present disclosure provides an electronic device, which adopts the following technical solution:

[0054] An electronic device comprises a memory, a processor, and a program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the multi-objective optimization planning method for deep peak-shaving transformation of thermal power units as described in the first aspect of the present disclosure are implemented.

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] This paper considers multiple economic factors, including total unit transformation cost, unit coal consumption cost, life loss cost, oil input fuel cost, environmental compensation cost, and peak-shaving compensation fee, while striving to maximize the deep peak-shaving capacity after unit transformation. A more comprehensive multi-objective optimization planning model for deep peak-shaving transformation of thermal power units is established, achieving the goal of maintaining optimal deep peak-shaving capacity to absorb as much renewable energy power generation as possible while maintaining economic efficiency.

[0057] This paper adopts a multi-objective particle swarm algorithm based on a periodic elite competition strategy. Based on the advantages of easy implementation and fast convergence of the multi-objective particle swarm algorithm, it selects elite particles to implement a competition mechanism regularly, so as to achieve the purpose of improving population diversity and preventing the algorithm from falling into local optimality.

[0058] Based on the solution results of the multi-objective model, this paper can obtain the optimal transformation plan for deep peak regulation of thermal power units in the power system to be transformed, meet the peak regulation needs through unit combination, and improve the renewable energy consumption situation. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] The accompanying drawings, which constitute a part of the present disclosure, are used to provide a further understanding of the present disclosure. The exemplary embodiments of the present disclosure and their descriptions are used to explain the present disclosure and do not constitute an improper limitation to the present disclosure.

[0060] Figure 1 This is a flow chart of a multi-objective optimization planning method for deep peak-shaving transformation of thermal power units in the first embodiment of the present disclosure;

[0061] Figure 2 This is a flow chart of a multi-objective particle swarm algorithm based on a periodic elite competition strategy in the first embodiment of the present disclosure;

[0062] Figure 3 is a schematic diagram of the competition strategy in the first embodiment of the present disclosure;

[0063] Figure 4 This is a structural diagram of a multi-objective optimization planning model for deep peak-shaving transformation of thermal power units in the first embodiment of the present disclosure;

[0064] Figure 5 Schematic diagram of the topology structure of the IEEE 39-node system in the first embodiment of the present disclosure;

[0065] Figure 6is a schematic diagram of the Pareto front curve in the first embodiment of the present disclosure;

[0066] Figure 7 It is a structural block diagram of the multi-objective optimization planning system for deep peak-shaving transformation of thermal power units in the second embodiment of the present disclosure. DETAILED DESCRIPTION

[0067] The present disclosure will be further described below with reference to the accompanying drawings and embodiments.

[0068] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of the present disclosure. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present disclosure belongs.

[0069] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present disclosure. As used herein, unless the context clearly indicates otherwise, the singular form is intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0070] In the absence of conflict, the embodiments of the present disclosure and the features thereof may be combined with each other.

[0071] Example 1

[0072] Embodiment 1 of the present disclosure introduces a multi-objective optimization planning method for deep peak-shaving transformation of thermal power units.

[0073] like Figure 1 A multi-objective optimization planning method for deep peak-shaving transformation of thermal power units is shown, including:

[0074] Obtain system parameters of thermal power units;

[0075] Based on the obtained system parameters of thermal power units, a multi-objective optimization planning model for deep peak-shaving transformation of thermal power units is constructed;

[0076] Solve the constructed multi-objective optimization planning model for deep peak-shaving transformation of thermal power units and obtain the optimization plan for deep peak-shaving transformation of thermal power units;

[0077] Among them, the constructed multi-objective optimization planning model for deep peak-shaving transformation of thermal power units includes an upper-level model and a lower-level model; the upper-level model takes the minimum total cost of deep peak-shaving and the maximum deep peak-shaving capacity as the upper-level objective function, and the minimum load of the thermal power unit after transformation and optimization as the decision variable; the lower-level model takes the minimum total operating cost of the unit as the lower-level objective function, and the unit output power and node voltage phase angle as the decision variables.

[0078] As one or more implementation methods, an upper-level model in a multi-objective optimization planning model for deep peak-shaving transformation of thermal power units is constructed, specifically:

[0079] First, the model decision variables are set. The decision variables of the upper model will be used as the deep peak-shaving transformation plan for the thermal power unit, so they are set to the minimum load of the unit after the transformation. The minimum load of the i-th unit after the transformation can be expressed as: P i，min .

[0080] Secondly, the model objectives are set. Objective 1 is set to minimize the total cost of deep peak regulation of the unit. The total cost of unit peak regulation needs to take into account the total cost of unit transformation. To facilitate planning and calculation, the annualized transformation cost is calculated as the total cost of unit transformation. The total cost of unit transformation can be expressed as:

[0081]

[0082] Where N is the total number of thermal power units in the power system to be transformed; S build is the unit deep peak load capacity transformation cost; r is the annual interest rate; y i is the life of the i-th unit; P i，mina is the minimum load of the i-th unit before transformation; P i，min is the minimum load of the i-th unit after transformation, when P i，min Equal to P i，mina When , it means that the i-th unit has not been modified.

[0083] The goal of minimizing the total cost of deep peak regulation of the unit includes two aspects. The total transformation cost of the unit is obtained by calculating the decision variables of the upper model, and the total operating cost of the unit is obtained from the lower model. The first goal can be expressed as:

[0084] minimize:F1=f1+f2

[0085] Among them, F1 represents the total cost of deep peak regulation of the unit; f1 represents the total transformation cost of the unit; f2 represents the total operating cost of the unit.

[0086] The second goal is to maximize the system's deep peak-shaving capacity, where the system's deep peak-shaving capacity refers to the difference between the sum of the minimum loads of all thermal power units in the system before the transformation and the sum of the minimum loads of all thermal power units in the system after the transformation. The second goal can be expressed as:

[0087]

[0088] Among them, P i，mina is the minimum load of unit i before transformation.

[0089] Finally, the model constraints are set, including the minimum load constraint of the unit after the transformation and the deep peak-shaving capacity constraint of the system after the transformation.

[0090] The minimum load constraint of the unit after transformation can be expressed as:

[0091] P i，minc ≤P i，min ≤P i，mina

[0092] Among them, P i，minc is the minimum load that the i-th unit can reach after the transformation.

[0093] The deep peak-shaving capacity constraint of the system after transformation can be expressed as:

[0094]

[0095] Among them, P bulild It represents the minimum demand for deep peak-shaving capacity of the power system to be transformed, provided by the power system dispatching department.

[0096] As one or more implementation methods, a lower-level model in a multi-objective optimization planning model for deep peak-shaving transformation of thermal power units is constructed, specifically:

[0097] First, the model decision variables are set. The lower model optimizes the total operating cost of the unit based on the decision variables based on the minimum load of the unit after transformation. Therefore, the decision variables are set as the unit output power and the node voltage phase angle. The DC power flow method is used to obtain the operating status of the power system, where the output power of the i-th unit in the t-th period can be expressed as P i，t , the voltage phase angle of the jth node in the tth period can be expressed as θ j，t .

[0098] Secondly, the model objective is set to minimize the total operating cost of the unit. The total operating cost of the unit takes into account many factors, including the unit's coal consumption cost, oil input fuel cost, life loss cost, environmental compensation cost and peak-shaving compensation cost.

[0099] The coal consumption of thermal power generation units can be fitted into the unit output power P i，t The quadratic coal consumption characteristic equation of the unit is given, so the coal consumption cost of the unit can be expressed as:

[0100] C a (P i，t )=S coal [a i (P i，t ) 2 +b i P i，t +c i ]

[0101] Among them, S coal Indicates the unit price of coal purchased by thermal power plants; a i、b i 、c i Represents the parameters of the coal consumption characteristic equation of the i-th unit.

[0102] The life loss cost of thermal power units is generally generated during the deep peak regulation stage. The i-th unit is running at the minimum load P before the transformation. i，mina The above period is the normal peak load regulation stage, and the operation is at the minimum load P after the transformation. i，min and the minimum load P before transformation i，mina The period between the two periods is the deep peak regulation stage. During the deep peak regulation stage, the thermal stress of the rotor shaft system caused by the low output of the thermal power unit will cause life loss. The life loss cost can be expressed as:

[0103] C b (P i，t )=S i，loss ×n(B i，t )

[0104] Among them, S i，loss represents the total construction cost of the i-th unit; B i，t represents the load rate of the i-th unit in the t-th period, B i，t =P i，t / P i，N , where P i，N represents the rated load of the i-th unit; n(B i，t ) is the deep peak load loss rate of thermal power units. The equation of deep peak load loss rate of thermal power units with respect to load rate is:

[0105]

[0106] Among them, d i 、e i 、f i 、g i Represents the deep peak regulation loss rate parameter of the i-th unit.

[0107] The fuel cost of thermal power units is generally generated during the peak load regulation phase of oil injection. The minimum non-oil injection load of the i-th unit is set to P i，minb The minimum load for oil injection is equal to the minimum load P that can be achieved after the transformation. i，minc , there is P i，minc <P i，minb The i-th unit is running at P i，minb With P i，mina The period between the two is the non-oil injection deep peak regulation stage. i，minc With P i，minb The period between the two periods is the peak-shaving stage of oil injection. When the thermal power unit is operating in the peak-shaving stage of oil injection, the coal combustion becomes unstable, and the boiler oil injection needs to be stabilized to ensure stable output. The fuel cost of oil injection can be expressed as:

[0108] C c (P i，t )=S oil ×p i，oil

[0109] Among them, S oil represents the unit oil price purchased by the thermal power plant; p i，oil Represents the oil injection per unit time of the i-th unit.

[0110] The power generation cost of the modified thermal power unit includes the unit coal consumption cost, oil input fuel cost and life loss cost, which are incurred in specific peak load regulation stages. The power generation cost of the unit can be expressed as:

[0111]

[0112] During the deep peak regulation phase of oil injection, the unit generates pollutants such as smoke and nitrogen oxides by injecting oil to maintain stable coal combustion. Environmental compensation costs need to be paid for pollution control. The environmental compensation costs can be expressed as:

[0113]

[0114] Among them, S env Indicates the environmental surcharge generated per unit of oil burned.

[0115] In order to fully stimulate the willingness of thermal power plants to participate in deep peak regulation, the peak regulation compensation fee is set in the peak regulation auxiliary service market model, and the minimum load P that can be achieved after the transformation is i，minc and the minimum load P before transformation i，mina The compensation fee is divided into a stepped range, and the peak load compensation fee can be expressed as:

[0116]

[0117] Among them, S i，t，offset It represents the compensation cost per unit of power generation of the i-th unit in the t-th period. The setting method is shown in Table 1, where P offset is the unit compensation cost ladder span value, S offset represents the upper limit of unit compensation cost, that is,

[0118]

[0119] Table 1 Unit Compensation Cost Level Table

[0120]

[0121] The lower model objective can be expressed as:

[0122]

[0123] Where T represents the optimization period of the lower model.

[0124] Finally, model constraints are set, including ramp constraints, unit output constraints, node voltage phase angle constraints, power balance constraints, and branch flow constraints.

[0125] The climbing constraint can be expressed as:

[0126]

[0127] Among them, P i，t-1 A represents the output power of the i-th unit in the t-1 period; i Indicates the maximum ramp rate of the i-th unit.

[0128] The unit output constraint can be expressed as:

[0129] P i，min ≤P i，t ≤P i，N

[0130] Among them, P i，min represents the minimum load of the i-th unit after transformation, P i，t represents the output power of the i-th unit in the t-th period, P i，N represents the rated load of the i-th unit.

[0131] The node voltage phase angle constraint can be expressed as:

[0132] θ j，min ≤θ j，t ≤θ j，max

[0133] Among them, θ j，min ,θ j，max They represent the minimum and maximum values of the node voltage phase angle of the jth node, θ j，t It represents the voltage phase angle of the j-th node in the t-th time period.

[0134] The power balance constraint can be expressed as:

[0135]

[0136] Among them, P j，t represents the output power of the jth node in the tth period; L j represents the load of the jth node; n represents the total number of nodes connected to the jth node; θ k，t and X k They represent the voltage phase angle of node k connected to the jth node and the line impedance between nodes respectively.

[0137] The branch power flow constraint can be expressed as:

[0138]

[0139] in, and They represent the maximum power flow limit and impedance value of the branch between the j1th node and the j2th node respectively; They represent the voltage phase angles of nodes j1 and j2 in the tth period respectively.

[0140] As one or more implementation methods, a multi-objective particle swarm algorithm based on a periodic elite competition strategy is set. Based on the advantages of easy implementation and rapid convergence of the multi-objective particle swarm algorithm, elite particles are selected to periodically implement the competition strategy to achieve the purpose of improving population diversity and preventing the algorithm from falling into local optimality. The algorithm flow chart is as follows Figure 2 As shown:

[0141] First, create a pop Set the current number of iterations to Iter, the maximum number of iterations to MaxIter, m targets and n decision variables, and initialize the positions and velocities of particles with n dimensions, where the velocity of particle k is expressed as:

[0142] V k (Iter)=(v k，1 (Iter), v k，2 (Iter),…,v k，n (Iter))

[0143] Position is represented by:

[0144] X k (Iter)=(x k，1 (Iter), x k，2 (Iter),…,x k，n (Iter))

[0145] The position range of particle k is expressed as:

[0146] (x kmin，1 ,…,x kmin，n )≤X k (Iter)≤(x kmax，1 ,…,x kmax，n )

[0147] In addition, an external archive needs to be established to save the positions of non-dominated solutions and individual optimal solutions and their corresponding positions during the algorithm operation, and then the iteration can be started.

[0148] Secondly, calculate N according to the objective function pop The fitness values of particles are calculated and sorted according to the dominance relationship. The fitness value of particle k is expressed as:

[0149] F k =(F k，1 (Iter), F k，2 (Iter),…,F k，m (Iter))

[0150] If k1 and k2 are two particles in the population, All fitness values are better than but Dominate It is called the dominated solution. If it is not dominated by any solution, it is called a non-dominated solution; and All fitness values in the solution are not completely better than each other, so they do not dominate each other. After completing the non-dominated sorting, the positions of the non-dominated solutions in the external archive are updated according to the sorting results. pop The fitness value of each particle updates the individual optimal position and the corresponding fitness value, and then sets the path decision amount c(Iter), which can be expressed as:

[0151]

[0152] When c(Iter) = 0, step (a) is executed to select the particle with the optimal solution as the leader particle to guide the rest of the particles to learn; when c(Iter) = 1, step (b) is executed to let the winner of the competition between the pairs of elite particles guide the rest of the particles to learn.

[0153] Step (a): Using the leader particle strategy, the non-dominated solution is used as the candidate solution set of the optimal solution, the optimal solution is randomly selected through the roulette strategy and the corresponding position is obtained from the external archive.

[0154] Based on the optimal solution position X best (Iter), for N pop The position and velocity of each particle are updated, and the individual optimal solution position of particle k is set to X k，best (Iter), the speed update formula is expressed as:

[0155] V k (Iter+1)=ω down ωV k (Iter)+c up1 c1(X best (Iter)-X k (Iter))+c up2 c2(X k，best (Iter)-X k (Iter))

[0156] Among them, V k (Iter+1) represents the updated velocity of particle k; ω represents the inertia coefficient; c1 represents the overall learning coefficient; c2 represents the individual learning coefficient; ω down Indicates the inertia attenuation coefficient, which gradually decays with the increase of the number of iterations, c up1 with c up2 It represents the learning acceleration coefficient, which gradually increases with the number of iterations. An adaptive strategy is used to ensure the global search capability in the early stage and accelerate the convergence speed in the later stage.

[0157] The position update formula is expressed as:

[0158] X k (Iter+1)=X k (Iter)+V k (Iter+1)

[0159] Among them, X k (Iter+1) represents the updated position of particle k.

[0160] Step (b): Adopting the elite competition strategy, the crowding distance sorting is performed on the basis of non-dominated sorting, and the elite particle set is determined according to the sorting result. Two particles are randomly selected from the elite particle set to compete to guide the learning of the particles to be updated.

[0161] In order to measure the quality of each solution in the same Pareto front, a crowding distance is assigned to each solution. Therefore, the obtained non-dominated solutions are sorted by crowding distance. Assuming that particle k corresponds to a non-dominated solution, the crowding distance is expressed as:

[0162]

[0163] Where dist(k) represents the crowding distance of particle k; f i max With f i min They represent the maximum and minimum fitness values under target i in the non-dominated solution; f i (k+1) and f i (k-1) represents the fitness value of the particles adjacent to particle k in the non-dominated solution under target i.

[0164] After performing non-dominated sorting and crowding distance sorting on the fitness value of each particle in the population, the elite particle set and the particle set to be updated are obtained. Then, two elite particles p and q are randomly selected to implement a competitive strategy for the particle v in the particle set to be updated, such as Figure 3 As shown, if the angle θ between particles q and v is qv Less than the angle θ between p and v pv, then particle q wins the competition and acts as the winner particle to guide particle v to update. The speed update formula is expressed as:

[0165] V v (Iter+1)=ωV v (Iter)+c3(X q (Iter)-X v (Iter))

[0166] Among them, V v (Iter+1) and V v (Iter) represents the speed of the particle v to be updated after and before the update; c3 represents the winner learning coefficient; X q (Iter) and X v (Iter) represents the positions of the winner particle q and the particle to be updated v before the update.

[0167] The position update formula is expressed as:

[0168] X v (Iter+1)=X v (Iter)+V v (Iter+1)

[0169] Among them, X v (Iter+1) represents the updated position of the particle v to be updated.

[0170] Repeat the random selection of two particles in the elite particle set and implement the competition strategy until all particles in the particle set to be updated are updated, and then the elite particles are updated at the original speed.

[0171] Finally, it is determined whether the current number of iterations Iter reaches the maximum number of iterations MaxIter. If the condition is not met, it returns to perform the Iter+1th iteration. If the condition is met, the optimization is completed.

[0172] As one or more implementation methods, according to the minimum load of the transformed unit provided by the upper model, the nonlinear programming problem of the lower model is solved based on the Gurobi solver, the total operating cost of the unit is optimized and returned to the upper model. The structure diagram of the two-level optimization model is as follows: Figure 4 shown.

[0173] As one or more implementation methods, a multi-objective particle swarm optimization algorithm based on a periodic elite competition strategy is applied to the upper model to solve the Pareto optimal solution set to obtain a deep peak-shaving transformation plan for the thermal power units.

[0174] The decision variables, objectives and constraints are configured in the selected algorithm, where the constraints are calculated using the formula ((x kmin，1 ,…,xkmin，n )≤X k (Iter)≤(x kmax，1 ,…,x kmax，n )). In each iteration, the decision variables before being updated in this iteration are passed to the lower model as the deep peak-shaving transformation plan of the thermal power unit to optimize the total unit operating cost f2 in the upper model's unit deep peak-shaving total cost minimum target, as shown in the formula (minimize: F1 = f1 + f2). The two-objective calculation results of the upper model are used as the fitness value to complete the iteration process, and finally the Pareto optimal solution set can be obtained as the optimization result of the two objectives of the upper model. The decision variables of each solution in the Pareto optimal solution set are the minimum load of the unit after the transformation, corresponding to a variety of deep peak-shaving transformation plans for thermal power units. Afterwards, the final transformation plan can be selected based on the focus of actual needs.

[0175] This embodiment takes the IEEE39 node standard calculation example system as an example to further illustrate the specific implementation process of the present invention. The IEEE39 node standard calculation example system contains a total of 10 power generation nodes, 18 load nodes and 46 branches. The 37, 38 and 39 power generation nodes in the IEEE39 node system are set as wind turbines, and the typical daily power generation data is used to provide peak load demand. The remaining power generation nodes are set as thermal power units to be transformed. The topology structure is as follows: Figure 5 shown.

[0176] In the upper model, the life of thermal power unit y i , rated power P i，N As shown in Table 2.

[0177] Table 2 Upper model parameters

[0178]

[0179] Minimum load P of the unit before transformation i，mina All rated power P i，N 50% of the minimum load P that the unit can reach after the transformation i，minc All rated power P i，N 20% of the total cost of deep peak load regulation, set the annual interest rate r = 0.06, and the unit deep peak load regulation capacity transformation cost S build = 500,000 yuan / MW, the minimum demand for deep peak load regulation capacity of the power system to be transformed P bulild =1000MW, and the capacity of the three wind turbines is 700MW each.

[0180] In the lower model, the total construction cost of the unit is S i，loss , oil injection per unit time p i，oil , Coal consumption characteristic equation parameter a i 、b i 、c iAs shown in Table 3.

[0181] Table 3 Lower layer model parameters

[0182]

[0183] Minimum load P of the unit without oil injection after transformation i，minb All rated power P i，N 35%, deep peak load loss rate parameter d i 、e i 、f i 、g i are set to -0.004, 0.007, -0.004, and 0.0008 respectively. The unit coal price S purchased by the thermal power plant coal = 1200 yuan / ton, unit oil price S oil = 9000 yuan / ton, environmental surcharge S generated by unit oil combustion env = 1000 yuan / ton, unit compensation cost S i，t，offset =1000 yuan / MW, the maximum ramp rate of the unit is A i are all set to 3%, the minimum and maximum values of the node voltage phase angle θ j，min ,θ j，max Set them to -360 and 360 respectively.

[0184] After the model parameters are set, the algorithm is set, where the maximum number of iterations MaxIter = 100, c(Iter) is set to 1 once after every 4 iterations, the number of iterations Iter = 1, and the population size N pop =50, n=7 is set according to the number of thermal power units to be transformed in the upper model, m=2 is set according to the target number of the upper model, the parameters of the Gurobi solver in the lower model are configured, the optimization period T=24, the unit time is set to 1 hour, and the output data of the three wind turbines are all used on a typical day. The multi-objective particle swarm algorithm based on the periodic elite competition strategy is used to solve the multi-objective optimization planning model of the deep peak-shaving transformation of thermal power units on a typical day. The Pareto front obtained is as follows Figure 6 shown.

[0185] This embodiment sets two overall goals: minimizing the total peak-shaving cost of thermal power units and maximizing the deep peak-shaving capacity. A two-layer optimization model is built, and reasonable constraints are set considering the actual situation. In order to obtain a better Pareto solution set, a multi-objective particle swarm algorithm based on a periodic elite competition strategy is used to solve the problem. The dynamic balance between the economy and flexibility of the deep peak-shaving transformation plan for thermal power units is achieved, and the optimal deep peak-shaving capacity is maintained under the premise of economy, so as to alleviate the peak-shaving pressure of the power grid and absorb as much new energy as possible.

[0186] Example 2

[0187] The second embodiment of the present disclosure introduces a multi-objective optimization planning system for deep peak-shaving transformation of thermal power units.

[0188] like Figure 7 A multi-objective optimization planning system for deep peak-shaving transformation of thermal power units is shown, comprising:

[0189] an acquisition module configured to acquire system parameters of a thermal power unit;

[0190] A modeling module is configured to construct a multi-objective optimization planning model for deep peak-shaving transformation of thermal power units based on the acquired thermal power unit system parameters;

[0191] an optimization module configured to solve the constructed multi-objective optimization planning model for deep peak-shaving transformation of thermal power units to obtain an optimization plan for deep peak-shaving transformation of thermal power units;

[0192] Among them, the constructed multi-objective optimization planning model for deep peak-shaving transformation of thermal power units includes an upper-level model and a lower-level model; the upper-level model takes the minimum total cost of deep peak-shaving and the maximum deep peak-shaving capacity as the upper-level objective function, and the minimum load of the thermal power unit after transformation and optimization as the decision variable; the lower-level model takes the minimum total operating cost of the unit as the lower-level objective function, and the unit output power and node voltage phase angle as the decision variables.

[0193] The detailed steps are the same as the multi-objective optimization planning method for deep peak-shaving transformation of thermal power units provided in Example 1, and will not be repeated here.

[0194] Example 3

[0195] A third embodiment of the present disclosure provides a computer-readable storage medium.

[0196] A computer-readable storage medium stores a program thereon, which, when executed by a processor, implements the steps of the multi-objective optimization planning method for deep peak-shaving transformation of thermal power units as described in the first embodiment of the present disclosure.

[0197] The detailed steps are the same as the multi-objective optimization planning method for deep peak-shaving transformation of thermal power units provided in Example 1, and will not be repeated here.

[0198] Example 4

[0199] A fourth embodiment of the present disclosure provides an electronic device.

[0200] An electronic device includes a memory, a processor, and a program stored in the memory and executable on the processor. When the processor executes the program, the steps of the multi-objective optimization planning method for deep peak-shaving transformation of thermal power units as described in the first embodiment of the present disclosure are implemented.

[0201] The detailed steps are the same as the multi-objective optimization planning method for deep peak-shaving transformation of thermal power units provided in Example 1, and will not be repeated here.

[0202] The foregoing description is merely a preferred embodiment of the present disclosure and is not intended to limit the present disclosure. Those skilled in the art will readily appreciate that various modifications and variations are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present disclosure shall be included within the scope of protection of the present disclosure.

[0203] Although the above describes the specific implementation methods of the present disclosure in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present disclosure. Those skilled in the art should understand that on the basis of the technical solution of the present disclosure, various modifications or variations that can be made by those skilled in the art without creative work are still within the scope of protection of the present disclosure.

Claims

1. A multi-objective optimization planning method for deep peak-shaving transformation of thermal power units, characterized in that: include: Obtain system parameters of thermal power units; Based on the obtained system parameters of thermal power units, a multi-objective optimization planning model for deep peak-shaving transformation of thermal power units is constructed; Solve the constructed multi-objective optimization planning model for deep peak-shaving transformation of thermal power units and obtain the optimization plan for deep peak-shaving transformation of thermal power units; The constructed multi-objective optimization planning model for deep peak-shaving transformation of thermal power units includes an upper-level model and a lower-level model; the upper-level model takes the minimum total cost of deep peak-shaving and the maximum capacity of deep peak-shaving as the upper-level objective function, and the minimum load of the thermal power unit after transformation optimization as the decision variable; the lower-level model takes the minimum total operating cost of the unit as the lower-level objective function, and takes the unit output power and node voltage phase angle as the decision variables; The total cost of deep peak regulation includes the total transformation cost of the unit and the total operating cost of the unit; wherein the total transformation cost f1 of the unit is Where N is the total number of thermal power units in the power system to be transformed; S build is the unit deep peak load capacity transformation cost; r is the annual interest rate; y i is the life of the i-th unit; P i,mina is the minimum load of the i-th unit before transformation; P i,min is the minimum load of the i-th unit after transformation, when P i,min Equal to P i,mina When , it means that the i-th unit has not been modified; The deep peak-shaving capacity is the difference between the sum of the minimum loads of all thermal power units in the system before the transformation and the sum of the minimum loads of all thermal power units in the system after the transformation, that is, the deep peak-shaving capacity F2 is The constraints of the upper objective function include the minimum load constraint of the unit after transformation and the deep peak-shaving capacity constraint of the system after transformation; The total operating cost of the unit includes the unit coal consumption cost, oil input fuel cost, life loss cost, environmental compensation cost and peak load compensation cost; The constraints of the lower objective function include ramp constraints, unit output constraints, node voltage phase angle constraints, power balance constraints and branch power flow constraints.

2. A multi-objective optimization planning method for deep peak-shaving transformation of thermal power units as described in claim 1, characterized in that: The obtained thermal power unit system parameters at least include the life of the thermal power unit, the minimum load of each unit, the operating cost of the unit and the coal consumption of the thermal power unit.

3. A multi-objective optimization planning method for deep peak-shaving transformation of thermal power units as claimed in claim 1, characterized in that: The minimum load constraint of the unit after the transformation is: P i,minc ≤P i,min ≤P i,mina Among them, P i,minc is the minimum load that the i-th unit can achieve after the transformation; The deep peak-shaving capacity constraint of the system after the transformation is: Among them, P bulild It represents the minimum demand for deep peak-shaving capacity of the power system to be transformed, provided by the power system dispatching department.

4. A multi-objective optimization planning method for deep peak-shaving transformation of thermal power units as claimed in claim 1, characterized in that: The climbing constraint is: Among them, P i,t-1 A represents the output power of the i-th unit in the t-1 period; i represents the maximum ramp rate of the i-th unit; The unit output constraint is: P i,min ≤P i,t ≤P i,N Among them, P i,min represents the minimum load of the i-th unit after transformation, P i,t represents the output power of the i-th unit in the t-th period, P i,N represents the rated load of the i-th unit; The node voltage phase angle constraint is: i j,min ≤θ j,t ≤θ j,max Among them, θ j,min ,θ j,max They represent the minimum and maximum values of the node voltage phase angle of the jth node, θ j,t represents the voltage phase angle of the jth node in the tth period; The power balance constraint is: Among them, P j,t represents the output power of the jth node in the tth period; L j represents the load of the jth node; n represents the total number of nodes connected to the jth node; θ k,t and X k They represent the voltage phase angle of node k connected to the jth node and the line impedance between nodes respectively; The branch power flow constraint is: in, and They represent the maximum power flow limit and impedance value of the branch between the j1th node and the j2th node respectively; They represent the voltage phase angles of nodes j1 and j2 in the tth period respectively.

5. A multi-objective optimization planning method for deep peak-shaving transformation of thermal power units as claimed in claim 1, characterized in that: In the process of solving the constructed multi-objective optimization planning model for deep peak-shaving transformation of thermal power units, a multi-objective particle swarm algorithm based on a periodic elite competition strategy is used to solve the upper model, and a nonlinear programming solution method is used to solve the lower model to obtain the Pareto optimal solution set and the deep peak-shaving transformation plan of the thermal power units.

6. A multi-objective optimization planning method for deep peak-shaving transformation of thermal power units as claimed in claim 5, characterized in that: According to the minimum load of the modified unit obtained by solving the upper model, the lower model is solved based on the nonlinear programming solution method to optimize the total operating cost of the unit, and the obtained total operating cost of the unit is fed back to the upper model.

7. A multi-objective optimization planning method for deep peak-shaving transformation of thermal power units as claimed in claim 6, characterized in that: According to the total operating cost of the unit obtained by solving the lower model, the upper model is solved using a multi-objective particle swarm algorithm based on a periodic elite competition strategy, the Pareto optimal solution set is solved, and the deep peak-shaving transformation plan of the thermal power unit is obtained.

8. A multi-objective optimization planning system for deep peak-shaving transformation of thermal power units, characterized by: include: an acquisition module configured to acquire system parameters of a thermal power unit; A modeling module is configured to construct a multi-objective optimization planning model for deep peak-shaving transformation of thermal power units based on the acquired thermal power unit system parameters; an optimization module configured to solve the constructed multi-objective optimization planning model for deep peak-shaving transformation of thermal power units to obtain an optimization plan for deep peak-shaving transformation of thermal power units; The constructed multi-objective optimization planning model for deep peak-shaving transformation of thermal power units includes an upper-level model and a lower-level model; the upper-level model takes the minimum total cost of deep peak-shaving and the maximum capacity of deep peak-shaving as the upper-level objective function, and the minimum load of the thermal power unit after transformation optimization as the decision variable; the lower-level model takes the minimum total operating cost of the unit as the lower-level objective function, and takes the unit output power and node voltage phase angle as the decision variables; The total cost of deep peak regulation includes the total transformation cost of the unit and the total operating cost of the unit; wherein the total transformation cost f1 of the unit is Where N is the total number of thermal power units in the power system to be transformed; S build is the unit deep peak load capacity transformation cost; r is the annual interest rate; y i is the life of the i-th unit; P i,mina is the minimum load of the i-th unit before transformation; P i,min is the minimum load of the i-th unit after transformation, when P i,min Equal to P i,mina When , it means that the i-th unit has not been modified; The deep peak-shaving capacity is the difference between the sum of the minimum loads of all thermal power units in the system before the transformation and the sum of the minimum loads of all thermal power units in the system after the transformation, that is, the deep peak-shaving capacity F2 is The constraints of the upper objective function include the minimum load constraint of the unit after transformation and the deep peak-shaving capacity constraint of the system after transformation; The total operating cost of the unit includes the unit coal consumption cost, oil input fuel cost, life loss cost, environmental compensation cost and peak load compensation cost; The constraints of the lower objective function include ramp constraints, unit output constraints, node voltage phase angle constraints, power balance constraints and branch power flow constraints.

9. A computer-readable storage medium having a program stored thereon, characterized in that: When the program is executed by a processor, the steps of the multi-objective optimization planning method for deep peak-shaving transformation of thermal power units as described in any one of claims 1 to 7 are implemented.

10. An electronic device comprising a memory, a processor, and a program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps in the multi-objective optimization planning method for deep peak-shaving transformation of thermal power units according to any one of claims 1 to 7 are implemented.

Images