Generation and transmission distribution robust planning method considering multi-energy complementation and direct current transmission

Abstract

The invention relates to a power generation and transmission distribution robust planning method considering multi-energy complementation and direct current transmission. The method comprises the following steps: constructing a wind power output scene probability distribution uncertainty set based on a conditional deep convolutional generative adversarial network method; obtaining a wind power output scene probability distribution uncertainty set constraint based on a wind power output scene probability distribution uncertainty set, and establishing a power generation and transmission distribution robust extension planning model considering multi-energy complementation and direct current transmission; constraint conditions adopted by the power generation and transmission distribution robust extension planning model comprise power generation and transmission equipment investment construction constraint, power system operation constraint and wind power output scene probability distribution uncertainty set constraint; and solving the power generation and transmission distribution robust extension planning model by adopting a column and constraint generation algorithm to obtain a power supply and power grid integrated optimal planning scheme considering the power generation and transmission characteristics. Compared with the prior art, multi-energy complementation and direct current transmission optimization can be considered, and meanwhile, the planning scheme has robustness and economical efficiency.

Description

Technical Field

[0001] The present invention belongs to the technical field of power system planning, and relates to a combined generation and transmission planning method, in particular to a distributionally robust planning method for combined generation and transmission considering multi-energy complementarity and DC power transmission. Background Art

[0002] In order to accelerate the construction of a new type of power system, it is urgent to establish and explore a power system planning system and development path that meet the requirements of safe and economic operation, power supply guarantee, and new energy consumption.

[0003] Combined generation and transmission planning refers to the process of considering and optimizing generation planning and transmission network planning as a whole in power system planning. This process aims to ensure the economy, security, and reliability of the power system while meeting the power demand of social and economic development. One of the keys to combined generation and transmission planning is to accurately characterize the uncertainty of new energy. Stochastic optimization methods are based on the probability distribution model that uncertainty factors obey, and construct a typical scenario set for optimization, but the calculation scale is often large and the solution complexity is high. Robust optimization methods describe the fluctuation of uncertain parameters by setting an uncertainty set, and aim to find the optimal planning scheme that meets the worst-case scenario, and the optimization results are often too conservative.

[0004] Existing technical research shows that in the process of new energy power system planning, considering multi-energy complementarity can improve the economy and reliability of the planning scheme and enhance the new energy consumption capacity. In addition, cross-regional power transmission through DC channels is also an effective way to promote new energy consumption and reduce energy curtailment. However, existing power system planning technologies mainly focus on considering the multi-energy complementarity effect and DC power transmission optimization in a single power source planning or grid planning, ignoring their potential in the process of combined generation and transmission planning.

[0005] Therefore, how to conduct combined generation and transmission planning for power systems by considering multi-energy complementarity and DC power transmission optimization in an uncertain environment is a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0006] The purpose of the present invention is to overcome the above-mentioned defects existing in the prior art and provide a distributionally robust planning method for combined generation and transmission considering multi-energy complementarity and DC power transmission, which can consider multi-energy complementarity and DC power transmission optimization, and at the same time make the planning scheme have both robustness and economy.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] A distributionally robust planning method for combined generation and transmission considering multi-energy complementarity and DC power transmission includes the following steps:

[0009] The conditional deep convolutional generative adversarial network method is adopted to generate wind power output scenarios reflecting actual statistical characteristics and distribution characteristics, and a corresponding probability distribution uncertainty set of wind power output scenarios is constructed;

[0010] Based on the probability distribution uncertainty set of the wind power output scenarios, the constraints of the probability distribution uncertainty set of the wind power output scenarios are obtained, and a transmission and distribution robust expansion planning model considering multi-energy complementarity and DC external transmission is established. The constraint conditions adopted by the transmission and distribution robust expansion planning model include the investment and construction constraints of transmission and distribution equipment, the operation constraints of the power system, and the constraints of the probability distribution uncertainty set of the wind power output scenarios. Among them, the operation constraints of the power system include the output constraints of conventional units, the operation constraints of wind farms, the operation constraints of pumped storage, the operation constraints of DC channels, and the line power flow constraints;

[0011] The column and constraint generation algorithm is used to solve the transmission and distribution robust expansion planning model, and the optimal integrated planning scheme of the power source and grid considering the transmission and distribution characteristics is obtained.

[0012] Furthermore, the probability distribution uncertainty of the wind power output scenarios is constrained by the 1-norm and ∞-norm sets to construct a probability distribution uncertainty set of the wind power output scenarios.

[0013] Furthermore, the transmission and distribution robust expansion planning model is a two-stage three-layer model. The first stage contains one layer to determine the transmission and distribution expansion planning scheme. The second stage contains two layers, the outer layer searches for the worst probability distribution fluctuation of the wind power output scenarios, and the inner layer determines the simulation operation scheme of the power system in each scenario.

[0014] Furthermore, the objective function of the transmission and distribution robust expansion planning model is constructed based on the planning investment cost and simulation operation cost converted to each year.

[0015] Furthermore, the investment and construction constraints of the transmission and distribution equipment include the construction constraints of transmission lines, the construction constraints of generating units, and the total investment cost constraints.

[0016] Furthermore, the operation constraints of the power system also include the node power balance constraints, the node voltage phase angle constraints, the load shedding constraints, and the spinning reserve constraints.

[0017] Furthermore, the DC external transmission load requirements are considered in the node power balance constraints.

[0018] Furthermore, the operation constraints of the DC channel include the upper and lower limits of the transmission power, the ramp rate constraints, the power up and down adjustment logic constraints, the adjustment times constraints, and the external transmission power constraints.

[0019] The present invention also provides a computer-readable storage medium, including one or more programs for execution by one or more processors of an electronic device, where the one or more programs include instructions for executing the power generation, transmission, and distribution robust planning method considering multi-energy complementarity and DC power transmission as described above.

[0020] The present invention also provides an electronic device, including one or more processors, a memory, and one or more programs stored in the memory, where the one or more programs include instructions for executing the power generation, transmission, and distribution robust planning method considering multi-energy complementarity and DC power transmission as described above.

[0021] Compared with the prior art, the present invention has the following beneficial effects:

[0022] 1. The present invention constructs a power generation, transmission, and distribution robust expansion planning model considering multi-energy complementarity and DC power transmission, and can obtain a combined power generation and transmission joint planning scheme with both economy and robustness. In the constraint conditions, the output operation optimization of conventional units, wind turbines, and pumped storage units at different times is considered simultaneously, that is, the resource endowments of chemical energy, wind energy, and water energy in different scenarios are fully utilized to achieve multi-energy complementarity under the optimal economic goal. A DC channel operation constraint is also established to constrain the DC power transmission and electricity quantity at different times based on requirements such as the safe operation and maximum regulation capacity of the DC channel, meeting the demand for cross-regional power consumption.

[0023] 2. The present invention collaboratively considers the power supply and grid planning, and the planning model considers the collaborative characteristics of power generation and transmission. Finally, a power supply and grid integrated decision-making scheme considering the characteristics of power generation and transmission is obtained, which can realize the adaptability consideration of power supply and grid in the planning stage, and the planning reliability is high.

[0024] 3. The present invention uses a conditional deep convolutional generative adversarial network to generate a large number of wind power output scenarios, which not only improves the stability of the network, but also can accurately analyze the correlation between the input data sampling points and local input information by using the convolutional network, improving the quality of the samples generated by the generator and accelerating the convergence speed. At the same time, it can effectively solve the problem of the deviation of the 1-norm and ∞-norm probability distributions caused by insufficient historical output data.

[0025] 4. The present invention depicts the power flow characteristics of the transmission network based on the DC power flow method, which can greatly reduce the model scale.

[0026] 5. The present invention uses a column and constraint generation algorithm that does not require duality and can be parallel computed to solve the power generation, transmission, and distribution robust planning model considering multi-energy complementarity and DC power transmission, which can effectively improve the solution speed and reduce the convex relaxation gap. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 is a schematic flow chart of the present invention;

[0028] Figure 2 This is the basic structural flowchart of the conditional deep convolutional generative adversarial network corresponding to the embodiment of the present invention;

[0029] Figure 3 This is a schematic diagram of the modified Graver-6 node example system in the embodiment of the present invention. Detailed implementation manners

[0030] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. This embodiment is implemented on the premise of the technical solution of the present invention, and gives detailed implementation manners and specific operation processes, but the protection scope of the present invention is not limited to the following embodiments.

[0031] The distributionally robust optimization method combines stochastic optimization and robust optimization methods, and uses a probability distribution uncertainty set to characterize uncertainty, overcoming the shortcomings of the above two methods. Among them, the data-driven distributionally robust optimization method has the advantages of simple solution and good economy because it can make full use of historical data rather than moment information to guide decision-making. Based on the above advantages, this embodiment provides a power generation and transmission distributionally robust planning method considering multi-energy complementarity and DC external transmission, as Figure 1 shown, including the following steps:

[0032] S101. Use the conditional deep convolutional generative adversarial network method to generate wind power output scenarios reflecting actual statistical characteristics and distribution characteristics, and construct a corresponding probability distribution uncertainty set of wind power output scenarios.

[0033] In this embodiment, the probability distribution uncertainty of the wind power output scenario is constrained by the 1-norm and ∞-norm sets to construct a probability distribution uncertainty set of the wind power output scenario, specifically a confidence set jointly constrained by the 1-norm and ∞-norm.

[0034] As Figure 2 shown, the conditional deep convolutional generative adversarial network method used in this embodiment is specifically:

[0035] Add specified conditional information G(z|y) to the generator, splice it with the random noise z, and then input it into the convolutional layer of the generator. After gradually refining the convolutional features extracted from the random noise z and the conditional information y, the generator outputs wind power output sample data G(z|y). Connect the conditional information y with the generated sample G(z|y) and the historical sample x respectively, and input them into the discriminator. By judging the similarity between the generated sample distribution and the real sample distribution and whether the generated sample satisfies the condition y, the discriminator outputs a judgment result.

[0036] The loss functions L G and L D in the generator G and the discriminator D are respectively expressed as:

[0037]

[0038] where: E(·) represents the expected value; D(·) represents the probability of being judged true by the discriminator; P z represents the probability distribution that the noise data z follows.

[0039] During the training process of the conditional deep convolutional generative adversarial network, the objective function of the min-max game model between the generator and the discriminator is shown in Equation (3).

[0040]

[0041] By introducing the Wasserstein distance to describe the distance between G(z|y) and x, the problem of difficult training and mode collapse can be solved. Then, the objective function of the conditional deep convolutional generative adversarial network is transformed into:

[0042]

[0043] where: is the penalty function of the gradient of the discriminator's loss function; λ is the penalty factor.

[0044] The construction of the probability distribution uncertainty set of the wind power scenario is specifically as follows:

[0045] The randomness of the probability distribution is characterized by a confidence set, as shown in Equation (5).

[0046]

[0047] where: represents the set of all probability distributions that the uncertainty factors may follow; represents the true probability distribution; represents the reference probability distribution obtained from historical data; represents the Euclidean distance between the true probability distribution and the reference probability distribution; ε represents the allowable limit of the probability deviation.

[0048] Specifically, the 1-norm and ∞-norm are used to characterize the fluctuation of the probability distribution, as shown in Equation (6) and Equation (7).

[0049]

[0050] and hold under the following confidence constraints:

[0051]

[0052] where: Pr{·} represents the probability that the inequality within {·} holds. Define respectively represent the confidence levels at which the two norm constraints hold. If the given confidence levels are ω1 and ω ∞ , the historical data scale U, and the typical scenario S, then γ1 and γ ∞ can be expressed as:

[0053]

[0054] In summary, the uncertainty set of the wind power scenario probability distribution can be obtained as shown in Equation (10).

[0055]

[0056] Introduce auxiliary variables and to linearly represent Equation (10) as shown in Equation (11).

[0057]

[0058] S102. Based on the uncertainty set of the wind power output scenario probability distribution, obtain the constraints of the wind power output scenario probability distribution uncertainty set, and establish a distributionally robust expansion planning model for power generation and transmission considering multi-energy complementarity and DC external transmission.

[0059] In this embodiment, the distributionally robust expansion planning model for power generation and transmission is a two-stage three-layer model. The first stage includes one layer (the first layer min), aiming to find the joint power generation and transmission planning scheme of the power system with the minimum annual investment cost. The second stage includes two layers (max-min) inside and outside. The outer layer max finds the most severe fluctuation of the wind power scenario probability distribution under the planning scheme given in the first stage, and the inner layer min aims to minimize the annual operation cost under the given planning scheme and the most severe wind power scenario probability distribution, and find the operation scheme of the power system in each scenario.

[0060] The objective function of this distributionally robust expansion planning model for power generation and transmission is constructed based on the discounted annual planning investment cost C INV and the simulated operation cost C OPE . Among them, C INV includes the equivalent annual investment construction costs of transmission lines and conventional generating units, and C OPE includes the operation fuel consumption costs of conventional generating units, the peak regulation costs of pumped storage power stations, the wind curtailment penalty costs, and the load shedding penalty costs. The specific calculation methods are shown in Equations (12)-(16).

[0061] minC = C INV + σC OPE (12)

[0062] C INV = C INV,L + C INV ,G (13)

[0063]

[0064] In the formula: σ represents the equivalent coefficient between the annual investment cost and the operation cost, which is taken as 365 in this paper; T L and T G respectively represent the economic service life of the transmission line and the generating unit; r is the discount rate; l, k, g, a, b, and i are the line corridor, transmission line, thermal power unit, pumped-storage power station, wind power station, and load node numbers respectively; represents the set of line corridor channels; and respectively represent the set of available lines and the set of existing lines in corridor channel l; x l,k represents the construction status of line k in corridor channel l, represents its equivalent annual investment cost; represents the set of generating units to be built; x g represents the construction status of generating unit g to be built, represents its equivalent annual investment cost; p s represents the probability of occurrence of the s-th planning scenario; S represents the total number of planning scenarios considered; T represents the time period in scenario s; and respectively represent the set of generating units, the set of pumped-storage units, the set of wind turbines, and the set of load nodes; represents the operation cost per unit power generation of generating unit g; represents the peak regulation cost of the pumped-storage power station; and respectively represent the unit curtailment of wind power and load shedding penalty costs; represents the output of generating unit g at time period t in scenario s; represents the peak regulation output of pumped-storage power station a at time period t in scenario s; represents the curtailment power of wind power station b at time period t in scenario s; represents the load shedding power at load node i at time period t in scenario s.

[0065] The constraint conditions adopted by this transmission and generation distribution robust expansion planning model include transmission and generation equipment investment and construction constraints, power system operation constraints, and wind power output scenario probability distribution uncertainty set constraints. Among them, the transmission and generation equipment investment and construction constraints include transmission line construction constraints, generating unit construction constraints, and total investment cost constraints; the power system operation constraints include conventional unit output constraints, wind farm operation constraints, pumped-storage operation constraints, DC channel operation constraints, line power flow constraints, node power balance constraints, node voltage phase angle constraints, load shedding constraints, and spinning reserve constraints. Specifically:

[0066] 1) Probability distribution uncertainty set constraint for wind power scenarios

[0067] The probability distribution uncertainty set constraint for wind power scenarios is shown in Equation (11).

[0068] 2) Investment and construction constraints for power transmission and generation equipment

[0069] The investment and construction constraints for power transmission and generation equipment include transmission line construction constraints, generator set construction constraints, and total investment cost constraints.

[0070] The construction status variables of the to-be-built lines and the existing lines need to satisfy the constraints shown in Equations (17) and (18) respectively. For the lines in the same corridor, the logical constraint shown in Equation (19) and the construction quantity constraint shown in Equation (20) need to be satisfied.

[0071]

[0072] In the formula: represents the maximum number of existing lines that can be stored in corridor l; N l,min and N l,max represent the minimum and maximum values of the number of transmission lines allowed to be built in corridor l respectively.

[0073] The construction status variables of the to-be-built generator sets and the already-built generator sets need to satisfy the constraints shown in Equations (21) and (22) respectively.

[0074]

[0075] In the formula: represents the set of existing generator sets.

[0076] The total investment cost constraint is:

[0077]

[0078] In the formula: represents the total investment cost expected and set by the planning decision maker.

[0079] 3) Power system operation constraints

[0080] The power system operation constraints include conventional generator set operation constraints, wind power plant operation constraints, pumped storage power plant operation constraints, DC channel operation constraints, node power balance constraints, node voltage phase angle constraints, power flow constraints, load shedding constraints, and spinning reserve constraints.

[0081] The operation constraints of conventional generator sets mainly consider the upper and lower limits of output and the ramping constraints, which are shown in Equations (24) and (25) respectively.

[0082]

[0083] In the formula: and respectively represent the lower and upper limits of the power of the conventional generating unit ; and respectively represent the upward and downward ramp rate limits of the conventional generating unit g within the unit time period Δt.

[0084] The operating constraints of the wind farm are:

[0085]

[0086] In the formula: represents the predicted output of the wind farm b at time t under scenario s; is the set maximum wind curtailment ratio.

[0087] The pumped-storage power station is divided into two states: power generation and pumping, and they cannot be carried out simultaneously during operation. For a conventional pumped-storage power station, its power generation power can vary continuously, while the pumping power cannot be adjusted to a constant value, and it satisfies the following constraint conditions.

[0088]

[0089] E a,min ≤E a,s,t ≤E a,max (31)

[0090] E a,s,T =E a,s,0 (32)

[0091]

[0092] In the formula: is the power generation power of the pumped-storage power station at time t under scenario s; is the pumping power of the pumped-storage power station at time t under scenario s; and are respectively the upper and lower limits of the power generation power of the pumped-storage power station η gen ; η gen and η pump are respectively the energy conversion coefficients under the power generation and pumping conditions; and are respectively the state variables of the power generation and pumping conditions of the pumped-storage power station E a,s,t at time t under scenario s; E a,s,t is the reservoir water storage of the pumped-storage power station E a,max at time t under scenario s, and its upper and lower limits are E a,max and E a,min ; E a,s,0 and E a,s,T are respectively the pumped-storage power station The reservoir storage corresponding to the initial and final moments within the scheduling period.

[0093] The operation constraints of the DC channel mainly include the upper and lower limits of the transmission power, the ramping constraint, the power upward and downward adjustment logic constraint, the adjustment times constraint, and the power transmission volume constraint.

[0094]

[0095] In the formula: is the power transmitted by the DC channel during period t under scenario s ; and are respectively the upper and lower limits of the capacity of the DC channel ; and are both binary variables, representing the state variables of the power upward and downward adjustment of the DC channel during period t under scenario s respectively; and are respectively the upper limits of the power upward and downward adjustment rates; Q c,s is the daily planned power transmission volume of the DC channel c.

[0096] The node power balance constraint is:

[0097]

[0098] In the formula: and respectively represent the set of conventional generating units, the set of pumped storage units, and the set of wind turbine units at the load node i; represents the load at node i during period t under scenario s; α(i) and β(i) respectively represent the sets of transmission corridors with i as the head node and the end node; f l,k,s,t represents the active power transmitted by the transmission line k in the channel l during period t under scenario s. The DC power transmission load requirements are considered in the above node power balance constraint.

[0099] The node voltage phase angle constraint is:

[0100] θ i,min ≤ θ i,s,t ≤ θ i,max (41)

[0101] In the formula: θ i,s,t represents the voltage phase angle of node i during period t under scenario s, θ i,min and θ i,max respectively represent its lower and upper limits.

[0102] The power flow constraint is:

[0103] f l,k,t = xl,k b l,k (θ i,t -θ j,t ) (42)

[0104] x l,k f l,k,min ≤f l,k,s,t ≤x l,k f l,k,max (43)

[0105] In the formula: b l,k represents the susceptance value of line f l,k,min in channel f l,k,min ; f l,k,min and f l,k,max represent the lower and upper limits of the active power transmitted by the channel line, respectively.

[0106] The big-M method can be used to transform the bilinear term in Equation (43) into a linear constraint, as shown in Equation (44).

[0107] -M(1 - x l,k ) ≤ f l,k,s,t -b l,k (θ i,s,t -θ j,s,t ) ≤ M(1 - x l,k ) (44)

[0108] The load shedding constraint is:

[0109]

[0110] In the formula: represents the maximum allowable load shedding ratio at node i.

[0111] The spinning reserve constraint is:

[0112]

[0113] In the formula: R s,t is the spinning reserve capacity requirement of the system at time t under scenario s. In this paper, this value is set to 15% of the load. For the convenience of subsequent description, the constructed two-stage distributionally robust planning model of power generation and transmission considering wind-storage complementarity and DC external transmission can be written in the matrix expression form as shown in Equation (47).

[0114]

[0115] In the formula: and represent the coefficient vectors in the objective function; x represents the variable related to the construction investment; y sVariables related to operation in different scenarios; A, B, D, E, F, d, g, h are parameter matrices in the constraint conditions, where ξ s is the predicted wind power value in the s-th scenario.

[0116] In the above two-stage three-layer model, the third-layer operation simulation model simultaneously considers the output operation optimization of conventional units, wind turbines, and pumped-storage units at different times, that is, it makes full use of the resource endowments of chemical energy, wind energy, and water energy in different scenarios to achieve multi-energy complementarity under the goal of optimal economy. For example, in the scenario of large-scale wind power generation, the low-cost consumption of wind power is considered first. If there is surplus electricity, it is stored through pumped storage; secondly, in the scenario of no wind output, pumped-storage power generation is considered to make up for the power and electricity balance. If the pumped-storage power generation is restricted, the thermal power generation with the highest cost is called to meet the power and electricity requirements. Finally, multi-energy complementarity and economic optimization in different time scenarios can be achieved.

[0117] In the above planning model, first, the DC external transmission load requirements are considered in the power balance constraints at each time to meet the cross-regional consumption demand of electricity; secondly, the operation constraints of the DC channel are established, mainly including the upper and lower limits of the transmission power, the ramp constraint, the power up and down adjustment logic constraint, the adjustment times constraint, and the external transmission electricity constraint. Based on the requirements of the safe operation and maximum regulation capacity of the DC channel, the DC external transmission power and electricity at different times are constrained, and the DC external transmission requirements are fully considered to further improve the reliability of the planning results.

[0118] S103. Solve the power transmission and distribution robust expansion planning model by using the column and constraint generation algorithm to obtain the optimal integrated power grid and power source planning scheme considering the characteristics of power transmission and distribution.

[0119] The column and constraint generation algorithm without duality and parallel computing used in this embodiment is specifically as follows:

[0120] Decompose the original problem into a master problem MP and a subproblem SP, generate the cutting plane of the subproblem, and gradually add the cutting plane and the variables of the subproblem to the master problem; alternately iterate and solve the master-subproblem until the difference between the upper and lower bounds of the solution of the original problem is less than a pre-given threshold, and the optimal solution of the original problem is considered to be obtained.

[0121] The master problem corresponds to the min problem in the first stage, seeks the optimal planning decision that satisfies the investment and construction constraints, and provides the lower bound value for the original model.

[0122]

[0123] In the formula: is the introduced auxiliary variable; R is the maximum number of iterations set for the algorithm.

[0124] For the power transmission and distribution joint planning scheme given in the r-th iteration The sub - problem finds the worst - case probability distribution and provides an upper - bound value for the original model.

[0125]

[0126] In the above formula (50), for different wind power output scenarios, there is no coupling relationship between the corresponding system operation decision variables. The probability distribution uncertainty set and the feasible set of operation decision variables are independent of each other. Therefore, the min operation and the Σ operation can be interchanged, and the sub - problem is decomposed into the following form:

[0127]

[0128] In summary, the solution process of the distribution - robust planning model is as follows:

[0129] 1) Initialize the lower bound LB = 0, the upper bound UB = +∞, R = 1, and obtain the reference scenario probability distribution based on the conditional deep convolutional adversarial generation network method

[0130] 2) Solve MP to obtain the optimal solution and update the lower - bound value

[0131] 3) For the given x * , first solve the inner - layer min problem in parallel, and then further solve the middle - layer max problem. If the SP problem has a feasible solution, obtain the worst - case scenario probability distribution and the optimal solution of the SP problem and update the upper - bound value If UB - LB ≤ ε, the iterative convergence requirement is met, and return the optimal planning scheme x * , otherwise update the MP while adding a new decision variable y s,R+1 to the MP, and adding the optimal cut set:

[0132]

[0133] If there is no feasible solution, add a new decision variable y s,R+1 to the MP, and add the feasible cut set:

[0134]

[0135] 4) Update R = R + 1 and return to 2).

[0136] In this embodiment, the power system to be planned shown in Figure 3 is used as a test case for analysis. Figure 3Schematic diagram of the modified Graver-6 node case test system. This case test system has a total of 15 transmission corridors, where the solid lines represent those that have been developed and the dashed lines represent those to be developed. Each corridor can accommodate a maximum of 3 transmission lines. Node 1 already has two conventional generating units of 90 MW and 60 MW, Node 3 already has one conventional generating unit of 120 MW, and conventional generating units are to be newly built at Nodes 2, 3, and 6. In addition, the converter at Node 1 conducts DC external transmission after commutation, a 30 MW pumped-storage power station is built at Node 2, and a 50 MW wind turbine is built at Node 4. The investment and operation parameters of the load, lines, and various types of generating units adopt the original standard data of this test case. The investment recovery period is 15 years and the discount rate is 5%. The penalty costs for unit wind curtailment and load shedding are set at $500 / (MW·h) and $3000 / (MW·h) respectively. 1000 wind power output scenarios are generated using the conditional deep convolutional adversarial generation network method, and finally 10 typical scenarios are obtained by clustering. The confidence levels γ1 and γ ∞ both take the value of 99%.

[0137] To verify the effectiveness of the two-stage distributionally robust planning method for power generation and transmission considering wind-pumped storage complementarity and DC external transmission, 4 planning methods are set for comparison, as shown in Table 1.

[0138] Table 1 Settings of planning methods

[0139]

[0140]

[0141] After optimization calculations, the five planning schemes and results obtained by the five methods are shown in Table 2.

[0142] Table 2 Comparison of planning results obtained from different cases

[0143]

[0144] Compared with Plan 1, Plan 2 has a more economical planning solution, with the annual comprehensive planning cost reduced by 171.2 million US dollars. Since Plan 1 does not consider the configuration of pumped-storage power stations, two additional 60MW conventional generating units need to be built at Node 2 to meet the load demand. In addition, Plan 1 lacks peak shaving means, and the wind power output cannot be fully absorbed, resulting in a certain amount of wind curtailment. In contrast, Plan 2 can store the low-valley wind power through the complementary advantages of wind power and pumped storage and adjust it to be sent out during the peak load period, greatly improving the operating economy. Compared with Plan 2, although the two planning solutions are the same, Plan 3 considers the adjustable DC transmission power on the basis of Plan 2 and can make full use of the conventional units at Node 6 with better economic benefits for power generation. By optimizing and adjusting the DC transmission power, the system operating cost is significantly reduced, and the annual comprehensive cost is reduced by 5.44%. Considering the planning results of Plan 1 to Plan 3, it can be seen that in the process of joint power generation and transmission planning, by configuring pumped-storage power stations to cooperate in power generation adjustment and optimizing the DC transmission power, the comprehensive system planning cost can be reduced to varying degrees, and at the same time, the absorption capacity of new energy can be improved.

[0145] It can be seen from the example that the planning solution considering multi-energy complementarity is more economical, with the annual comprehensive planning cost reduced by 171.2 million US dollars. Since the non-multi-energy complementarity plan does not consider the configuration of pumped-storage power stations, two additional 60MW conventional generating units need to be built to meet the load demand. In addition, without considering multi-energy complementarity, the peak shaving means are relatively lacking, and the wind power output cannot be fully absorbed, resulting in a certain amount of wind curtailment. Considering multi-energy complementarity can store the low-valley wind power through the complementary advantages of wind power and pumped storage and adjust it to be sent out during the peak load period, greatly improving the operating economy.

[0146] The annual comprehensive planning cost of the data-driven distributionally robust optimization method is higher than that of the stochastic optimization method and lower than that of the robust optimization method. The stochastic optimization method makes planning decisions based on the given wind power output scenarios. The resulting planning solution has a relatively aggressive investment in conventional generating units and transmission lines, with a relatively weak grid structure. Although it has the best economy among the three planning solutions, its risk resistance is weak. The robust optimization method considers planning under the worst wind power output scenarios, which is prone to redundant equipment investment. However, the probability of extremely bad scenarios is usually small, resulting in low equipment utilization. Therefore, the planning solution obtained by the robust optimization method has strong robustness but poor economy. Compared with the planning solution obtained by the stochastic optimization method, the annual comprehensive planning cost increases by 4.23%. The data-driven distributionally robust optimization method considers optimizing the plan under the probability distribution of the worst wind power output scenarios. It can not only utilize the easily accessible information in the historical data of wind power output but also consider the uncertainty characteristics of the probability distribution of wind power output. The resulting planning solution achieves a balance between economy and risk resistance.

[0147] Tables 3 and 4 compare the planning results using the comprehensive norm and the single norm at different confidence levels. Among them, Tables 3 and 4 respectively keep the confidence levels of the 1-norm and ∞-norm constraints at 0.9. It can be seen that the annual comprehensive cost of the planning scheme obtained using the comprehensive norm is lower than that using the single norm, which indicates that using the comprehensive norm constraint can jointly describe the probability distribution of the wind power output scenarios from both the overall fluctuation degree and the maximum fluctuation degree, and can effectively reduce the conservatism of the planning scheme. Notice that when γ1 = 0.99 and γ ∞ = 0.9, the annual planning comprehensive costs obtained by only considering the ∞-norm constraint and considering the comprehensive norm constraint are the same, that is, at this time, the 1-norm has a smaller effect on the probability distribution uncertainty set, while the ∞-norm constraint has a greater impact on the uncertainty set.

[0148] Comparison of planning results under the comprehensive norm and 1-norm constraints in Table 3

[0149]

[0150]

[0151] Comparison of planning results under the comprehensive norm and ∞-norm constraints in Table 4

[0152]

[0153] Further study the influence of the historical data scale for constructing the probability distribution uncertainty set of wind power output scenarios on the planning results. Keeping the number of typical scenarios and the confidence level unchanged, use the conditional deep convolutional generative adversarial network method to generate wind power scenarios data of different scales, and finally all are clustered to obtain the same number of typical wind power output scenarios. Through simulation calculations, the obtained planning results are shown in Table 5. It can be seen from the table that as the scale of the wind power output scenario data considered increases, the annual comprehensive cost of the planning scheme gradually decreases. This indicates that as the amount of historical data information increases, the reference probability distribution of the wind power output scenarios is closer to the true probability distribution, the feasible region of the probability distribution of the max layer in the planning model becomes smaller, and at the same time, the deviation of the scenario probability distribution in the worst case found is reduced, which is also the significance of using the data-driven method.

[0154] Comparison of planning results under different generated scenario scales in Table 5

[0155]

[0156] When the above method is implemented in the form of software functional units and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs.

[0157] In another embodiment, an electronic device is provided, including one or more processors, a memory, and one or more programs stored in the memory. The one or more programs include instructions for executing the power generation, transmission, and distribution robust planning method considering multi-energy complementarity and DC power transmission as described above.

[0158] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) containing computer-usable program codes.

[0159] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0160] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to work in a particular manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction device that implements the functions specified in one or more of the processes Figure 1 or blocks Figure 1 specified in one or more of the processes and / or blocks.

[0161] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one or more of the processes Figure 1 or blocks Figure 1 specified in one or more of the processes and / or blocks.

[0162] The preferred specific embodiments of the present invention have been described in detail above. It should be understood that those of ordinary skill in the art can make many modifications and variations based on the concept of the present invention without creative work. Therefore, all technical solutions that can be obtained by those skilled in the art in the technical field based on the concept of the present invention through logical analysis, reasoning, or limited experiments on the basis of the prior art shall fall within the protection scope determined by the claims.

Claims

1. A method for planning power generation and transmission distribution considering multi-energy complementarity and DC transmission, characterized in that: The following steps are involved: The conditional deep convolutional generative adversarial network method is used to generate wind power output scenarios that reflect actual statistical characteristics and distribution characteristics, and the corresponding wind power output scenario probability distribution uncertainty set is constructed; Based on the uncertain set of probability distribution of wind power output scenarios, the uncertain set constraints of probability distribution of wind power output scenarios are obtained, and a power generation and transmission distribution blue stick expansion planning model considering multi-energy complementarity and DC transmission is established. The constraints adopted by the power generation and transmission distribution blue stick expansion planning model include power generation and transmission equipment investment and construction constraints, power system operation constraints and wind power output scenario probability distribution uncertain set constraints, wherein the power system operation constraints include conventional unit output constraints, wind farm operation constraints, pumped storage operation constraints, DC channel operation constraints and line flow constraints; The column and constraint generation algorithm is used to solve the generation and transmission distribution robust expansion planning model, and the optimal planning scheme for power supply and grid integration considering the generation and transmission characteristics is obtained.

2. The method for robust planning of power generation and transmission distribution considering multi-energy complementarity and DC transmission according to claim 1 is characterized in that: The uncertainty of the probability distribution of the wind power output scenario is constrained by a 1-norm and an ∞-norm set, and an uncertainty set of the probability distribution of the wind power output scenario is constructed.

3. The method for robust planning of power generation and transmission distribution considering multi-energy complementarity and DC transmission according to claim 1 is characterized in that: The power generation and transmission distribution expansion planning model is a two-stage three-layer model. The first stage includes a layer to determine the power generation and transmission expansion planning scheme. The second stage includes two layers, the outer layer looks for the worst probability distribution fluctuation of wind power output scenario, and the inner layer determines the simulated operation scheme of the power system in each scenario.

4. The method for robust planning of power generation and transmission distribution considering multi-energy complementarity and DC transmission according to claim 1 is characterized in that: The objective function of the generation and transmission distribution expansion planning model is constructed based on the planned investment costs and simulated operating costs converted to annual amounts.

5. The method for robust planning of power generation and transmission distribution considering multi-energy complementarity and DC transmission according to claim 1 is characterized in that: The power generation and transmission equipment investment and construction constraints include power transmission line construction constraints, power generation unit construction constraints and total investment cost constraints.

6. The method for robust planning of power generation and transmission distribution considering multi-energy complementarity and DC transmission according to claim 1 is characterized in that: The power system operation constraints also include node power balance constraints, node voltage phase angle constraints, load shedding constraints and spinning reserve constraints.

7. The method for robust planning of power generation and transmission distribution considering multi-energy complementarity and DC transmission according to claim 6 is characterized in that: The node power balance constraint takes into account the DC transmission load requirement.

8. The method for robust planning of power generation and transmission considering multi-energy complementarity and DC transmission according to claim 1 is characterized in that: The DC channel operation constraints include upper and lower limit constraints of transmission power, ramp constraints, power upward and downward adjustment logic constraints, adjustment number constraints and external power transmission constraints.

9. A computer-readable storage medium, characterized in that: It includes one or more programs for execution by one or more processors of an electronic device, and the one or more programs include instructions for executing a robust planning method for power generation and transmission distribution taking into account multi-energy complementarity and direct current transmission as described in any one of claims 1-8.

10. An electronic device, characterized in that: It includes one or more processors, a memory and one or more programs stored in the memory, and the one or more programs include instructions for executing a robust planning method for power generation and transmission distribution considering multi-energy complementarity and direct current transmission as described in any one of claims 1-8.

Images