Optimization method for joint maintenance of cascade hydropower-wind-thermal units considering uncertainty

Abstract

The present invention discloses a combined maintenance optimization method for cascade hydropower-wind-fired units taking uncertainty into account, comprising constructing a combined low-carbon maintenance optimization model for cascade hydropower-wind-fired units, solving the model to obtain a basic maintenance plan and a basic cost; solving the probability distribution function of the wind speed based on wind speed data and combining historical wind speed data at historical moments; obtaining the probability distribution function of wind power output according to the probability distribution function of the wind speed, and solving the wind power output range and the minimum deviation coefficient under different confidence levels; constructing a combined low-carbon maintenance optimization model for cascade hydropower-wind-fired units based on a time-varying CC-IGDT based on the basic cost and the minimum deviation coefficient and taking the wind power output range under different confidence levels as the uncertainty range; the present invention can effectively suppress the intermittent nature of wind power output, ensure the safety and stability of the system, and effectively improve the efficiency of solving uncertainty problems.

Description

Technical Field

[0001] The present invention relates to the field of electric power technology, and in particular to a combined maintenance optimization method for cascade hydropower-wind-thermal units taking uncertainty into account. Background Art

[0002] For conventional hydro-thermal power units, the maintenance plan is to arrange the maintenance interval under a series of optimization goals and system operation constraints, or to formulate the maintenance plan after deducing the fault law and evaluating the remaining life based on the full life cycle theory. For new energy units such as wind turbines, it is usually expected to arrange maintenance during the period of predicted low output and external environment friendly state, and these key factors that determine the quality of maintenance plans are often highly uncertain. In addition, wind farms generally contain a large number of wind turbines with the same installed capacity. How to coordinate the joint maintenance optimization of various units in large-scale wind farms has become a difficult problem. In addition, in the power system with a high proportion of new energy access, the randomness and volatility of wind power output will inevitably affect power dispatching, destroy system stability, reduce the balance of the power balance process over a long period of time, and thus interfere with the effective formulation of maintenance plans. How to effectively smooth the intermittent nature of wind power output, while maximizing the consumption of wind power, reasonably and accurately formulate a joint low-carbon maintenance plan for multiple types of units that adapt to uncertain disturbances is a difficult problem that needs to be solved in the current maintenance optimization field.

[0003] For conventional hydro-thermal power units, the optimization methods of unit maintenance change with the changes of diversified objectives and diversified application scenarios. At present, the modeling of generator unit maintenance problems at home and abroad is mainly divided into three categories: unit maintenance models with system operation reliability as the optimization target, system economy as the optimization target, and unit maintenance models that take into account multiple optimization targets. However, the optimization objects of the maintenance models described in the existing literature are mostly single thermal power or hydropower units, and the joint maintenance optimization of multiple types of units is rarely considered, and the target requirements such as the low carbon nature of the power system under the new situation are ignored. The southwest region of my country is rich in hydropower resources. At present, relatively complete and mature cascade hydropower station projects have been built, and the construction of other cascade power stations is also being continuously promoted. In order to actively respond to the "West-to-East Power Transmission" strategy to maximize the use of clean energy, it is necessary to construct a flexible maintenance optimization model for cascade hydro-thermal power units.

[0004] For wind turbines, most of the research on planned maintenance focuses on the formulation of maintenance strategies. Some scholars have carefully studied the hierarchical structure and operation mechanism of each component of the wind turbine, and are committed to achieving integrated optimal maintenance from components to the whole. Some scholars also believe that the maintenance order and fault correlation of each component can be ignored, and the entire wind turbine can be used as the optimization object, and the model can be constructed with the optimal maintenance interval as the core. In fact, the external environment of each unit in the same wind farm is not much different, the unit structure is similar, and large-scale power grids often arrange day-ahead dispatching based on wind farms. From the perspective of unit maintenance serving the operation of the large power grid, the wind farm can be considered as a separate individual to coordinate the joint maintenance optimization of wind turbines and other units.

[0005] For uncertainty optimization problems, the processing methods are usually robust optimization and stochastic programming. Studies have shown that the robust optimization method is based on the premise that all elements of the set meet the model constraints, and increases unnecessary economic losses to ensure the high reliability requirements of the system, and the decision results are too conservative. The stochastic programming method has a high dependence on historical data, and the decision effect is generally inferior to the worst effect borne by investors. In this scenario, the Information Gap Decision Theory (IGDT) that effectively handles uncertainty came into being. Different from the robust optimization model, IGDT aims to study the impact of uncertainty parameters on optimization problems based on preset goals, and the optimization target is the uncertainty parameter. Unlike the stochastic programming method, IGDT does not require a large amount of raw data and accurate probability distribution as support. At present, IGDT has been successfully applied in power trading, economic dispatch, power flow optimization, distribution network operation, etc. However, existing research rarely improves and optimizes the IGDT method. Summary of the invention

[0006] The purpose of the present invention is to overcome the deficiencies of the above-mentioned prior art and to provide a method for optimizing the joint maintenance of cascade hydropower-wind-thermal units taking into account uncertainty, so as to effectively smooth out the intermittent nature of wind power output, ensure the safety and stability of the system, and effectively improve the efficiency of solving uncertainty problems.

[0007] To achieve the above object, the technical solution of the present invention is:

[0008] A combined maintenance optimization method for cascade hydropower-wind-thermal units taking into account uncertainty, comprising:

[0009] Construct a low-carbon maintenance optimization model for cascaded hydropower-wind-thermal units, and solve the model to obtain the basic maintenance plan and basic cost;

[0010] Based on the wind speed data and combined with the historical wind speed data at historical moments, the probability distribution function of wind speed is solved;

[0011] The probability distribution function of wind power output is obtained according to the probability distribution function of wind speed, and the wind power output range and minimum deviation coefficient under different confidence levels are solved;

[0012] Based on the basic cost and the minimum deviation coefficient, and taking the wind power output range under different confidence levels as the uncertainty range, a joint low-carbon maintenance optimization model of cascade hydropower-wind-thermal units based on time-varying CC-IGDT is constructed.

[0013] The combined low-carbon maintenance optimization model of the cascade hydropower-wind-thermal units based on time-varying CC-IGDT is solved to obtain the total cost and the combined low-carbon maintenance optimization plan of the cascade hydropower-wind-thermal units.

[0014] Compared with the prior art, the present invention has the following beneficial effects:

[0015] The cascade hydropower-wind-fired unit joint low-carbon maintenance optimization model based on time-varying CC-IGDT constructed by the present invention is based on basic cost and minimum deviation coefficient, and is constructed with wind power output intervals under different confidence levels as uncertainty intervals, realizing an optimization process based on information support and an optimal opportunity constraint based on probability. The model fully considers the impact of wind power output uncertainty on unit maintenance, and can obtain a maintenance plan that adapts to uncertainty disturbances. The time-varying CC-IGDT can still seek a more economical maintenance plan after taking into account the uncertainty of wind power output. The maintenance optimization process fully considers the priority absorption of wind power characteristics and system operation characteristics, effectively smoothes the intermittent nature of wind power output, and ensures the safety and stability of the system; compared with the conventional CC-IGDT method, the decision-making conservatism is reduced, and a cascade hydropower-wind-fired joint maintenance plan with a cost lower than the expected cost is solved; compared with the scenario method, the optimization based on information support is realized, and the efficiency of solving uncertainty problems is effectively improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A flow chart of a combined maintenance optimization method for cascade hydropower-wind-thermal units taking into account uncertainty provided by an embodiment of the present invention;

[0017] Figure 2 is the wind speed probability confidence interval;

[0018] Figure 3 It is the optimization area of ​​time-varying CC-IGDT;

[0019] Figure 4 It is the optimization area of ​​time-varying CC-IGDT. DETAILED DESCRIPTION

[0020] Example:

[0021] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0022] See also Figure 1 As shown, it is a flow chart of the combined maintenance optimization method of cascade hydropower-wind-thermal units taking into account uncertainty provided in this embodiment, which mainly includes the following steps:

[0023] 101. Construct a low-carbon maintenance optimization model for cascaded hydropower-wind-thermal units, and solve the model to obtain the basic maintenance plan and basic cost.

[0024] In this step, based on the characteristics of fan maintenance that are easily affected by external factors, frequent failures, and immediate inspection, the following assumptions are set based on the principle of simplified model:

[0025] (1) Considering the similar system characteristics and external meteorological characteristics of wind turbines in the same region, wind power and hydropower units jointly participate in maintenance optimization;

[0026] (2) Generally, the maintenance cycle of a fan is short. With reference to a large number of literature, this embodiment sets the maintenance period to 3 days;

[0027] (3) During the maintenance period, the maintenance capacity value of a certain wind farm must be a fixed value. In this embodiment, the minimum value within the maintenance interval is taken.

[0028] 102. Based on the wind speed data and combined with the historical wind speed data at historical moments, the probability distribution function of wind speed is solved.

[0029] 103. The probability distribution function of wind power output is obtained according to the probability distribution function of wind speed, and the wind power output range and minimum deviation coefficient under different confidence levels are solved.

[0030] 104. Based on the basic cost and the minimum deviation coefficient, and taking the wind power output range under different confidence levels as the uncertainty range, a joint low-carbon maintenance optimization model of cascade hydropower-wind-thermal units based on time-varying CC-IGDT is constructed.

[0031] Since the cascade hydropower-wind-fired unit joint low-carbon maintenance optimization model based on time-varying CC-IGDT constructed in this step is based on the basic cost and the minimum deviation coefficient, and is constructed with the wind power output range under different confidence levels as the uncertainty range, it realizes the optimization process based on information support and the optimal opportunity constraint based on probability. The model fully considers the impact of wind power output uncertainty on unit maintenance, and can obtain a maintenance plan that adapts to uncertainty disturbances. The time-varying CC-IGDT can still seek a more economical maintenance plan after taking into account the uncertainty of wind power output. The maintenance optimization process fully considers the priority absorption of wind power characteristics and system operation characteristics, effectively smoothes the intermittent nature of wind power output, and ensures the safety and stability of the system; compared with the conventional CC-IGDT method, it reduces the conservatism of decision-making, and solves the cascade hydropower-wind-fire joint maintenance plan that is lower than expected; compared with the scenario method, it realizes the optimization based on information support, and effectively improves the efficiency of solving uncertainty problems.

[0032] 105. Solve the time-varying CC-IGDT-based cascade hydropower-wind-thermal unit joint low-carbon maintenance optimization model to obtain the total cost and the cascade hydropower-wind-thermal unit joint low-carbon maintenance optimization plan.

[0033] In a specific embodiment, the cascade hydropower-wind-thermal unit joint low-carbon maintenance optimization model is constructed in the following manner:

[0034] Ignoring the uncertainty of wind power, with the goal of minimizing economic losses and assisting the low-carbon operation of the power system, the unit maintenance cost, operating cost, and carbon emission cost targets are comprehensively considered, and an optimization target is constructed by the weighted summation of various costs, in which each weight coefficient can be used to systematize the penalty intensity of the cost value.

[0035] Construct the overall optimization target minC:

[0036]

[0037] Unit maintenance cost It is expressed as:

[0038]

[0039] Unit operating cost It is expressed as:

[0040]

[0041] Carbon emissions costs It is expressed as:

[0042]

[0043] Wherein, T is the maintenance period (in this embodiment, the maintenance period is one month, T=30); i is the unit number; Z is the total number of hydropower units, Z=Th+Tm, Th and Tm are the number of thermal power units and hydropower units respectively; W is the total number of wind farms; C jx,i,t , C jx,w,t They represent the maintenance costs of unit capacity hydro-thermal power units and fans respectively. Due to the limitation of human and material resources in actual situations, this cost varies with time within a month. When human and material resources are scarce, its value is higher; i,max is the maximum output of the hydro-thermal power unit; a i (t) is the maintenance 0-1 variable of unit i; a i (t) = 1 means that unit i is under maintenance at time t, a i (t) = 0 means that unit i is in a non-maintenance state at time t. w (t) is the maintenance continuous variable of the wind farm w, and its value reflects the capacity value that actually needs maintenance in the wind farm; P i (t) is the output of thermal power unit i at time t; c 0,i 、c 1,i 、c 2,i is the coal consumption coefficient of thermal power units; is the carbon cost per kilowatt-hour, which is 0.019 yuan / (kW·h), δ t is the time interval, i.e. the number of hours per day (the maintenance model of the present invention is based on days), which is 24 hours; They are the weights of unit maintenance cost, operating cost and carbon emission cost, respectively, and are set according to the actual carbon penalty requirements of the system.

[0044] The optimization process comprehensively considers the cascade hydroelectric coupling constraints, unit maintenance constraints, system operation control constraints, etc., and formulates a reasonable joint maintenance optimization strategy for the cascade hydropower-wind-thermal units, specifically:

[0045] (I) Cascade hydroelectric coupling constraints

[0046] (1) Hydropower station reservoir capacity balance equation

[0047] For any hydropower station k, its reservoir capacity is not only affected by its natural water flow, power generation flow, and water abandonment flow, but also by the power generation flow and water abandonment flow of the previous hydropower station k-1, which can be specifically expressed as:

[0048] v k,t =v k,t-1 +(q K,k-1,t +s k-1,t +n k,t -q K,k,t -s k,t )Δt (5)

[0049] In the formula, vk,t 、v k,t-1 are the storage capacities of hydropower station k at time t and time t-1 respectively; n k,t ,q K,k,t 、s k,t are the natural water flow (including rainfall, snowmelt, etc.), power generation flow, and abandoned water flow of hydropower station k at time t; K,k-1,t 、s k-1,t are respectively the power generation flow and abandoned water flow of hydropower station k-1 at time t.

[0050] For the cascade leading power station, the storage capacity balance equation is corrected to:

[0051] v 1,t =v 1,t-1 +(n 1,t -q K,1,t -s 1,t )Δt (6)

[0052] (2) Hydropower station output characteristic equation

[0053] P k (t) = η k q K,k,t (7)

[0054] Where P k (t) is the output of hydropower station k; η k is the flow-power conversion coefficient of hydropower station k, which can be obtained by fitting the historical power generation power and power generation flow of hydropower station k.

[0055] (3) Hydropower station output equation

[0056] For any hydropower station k, its output is equal to the sum of the outputs of all units in the hydropower station:

[0057]

[0058] Where P i (t) is the output of hydropower unit i, Ω k is the set of hydropower units in hydropower station k.

[0059] (4) Upper and lower limits of hydropower unit output

[0060] At any time, the output of the hydropower unit is between the minimum technical output and the maximum technical output:

[0061] (1-a i (t))P i,min ≤P i (t)≤(1-a i (t))P i,max (9)

[0062] Where P i,min , P i,max are the lower and upper limits of the output of hydropower unit i, respectively, i∈Ω Tm ; a i (t) is the 0-1 variable for the maintenance of the hydropower unit, a i (t) = 1 means that the hydropower unit i is under maintenance, a i (t) = 0 means that unit i is in a non-maintenance state.

[0063] (5) Upper and lower limits of hydropower station storage capacity

[0064] v k,min ≤v k,t ≤v k,max (10)

[0065] In the formula, v k,min 、v k,max are the lower and upper limits of the storage capacity of reservoir k respectively.

[0066] (6) Upper and lower limits of hydropower station power flow

[0067] Limited by the upper and lower limits of the hydropower unit output, the power generation flow of the hydropower station should be maintained within a certain range, namely:

[0068] q K,k,min ≤q K,k,t ≤q K,k,max (11)

[0069] In the formula, q K,k,min ,q K,k,max are the lower and upper limits of the power generation flow of reservoir k respectively.

[0070] (7) Constraints on the amount of water abandoned

[0071] In order to ensure the safe operation of the hydropower station, some water needs to be discharged at certain times to ensure that the reservoir capacity does not exceed the maximum allowable reservoir capacity, that is:

[0072]

[0073] (8) Ecological flow constraints

[0074] In order to ensure the ecological service function of rivers and maintain or restore the basic structure and function of river, lake and reservoir ecosystems, rivers need to have a minimum flow, namely:

[0075] q K,k,t +s k,t ≥ecoflow k (13)

[0076] Where, ecoflow krepresents the ecological flow demand of hydropower station k, and the present invention adopts the Tennant method to determine this value. This formula shows that hydropower station k needs to discharge at least the amount of water that meets the ecological flow in time period t.

[0077] (II) Unit maintenance constraints

[0078] (1) Maintenance continuity constraints

[0079] For any conventional hydro-thermal power unit i, the maintenance plan has time continuity, that is, maintenance items cannot be suspended within the maintenance cycle:

[0080]

[0081] (2) Maintenance period constraints

[0082] For any conventional hydro-thermal power unit i, the time consumed for maintenance is fixed, that is:

[0083]

[0084] Where, T cx,i It is the preset maintenance duration based on the installed capacity of unit i, historical maintenance rules, etc., that is, the maintenance period.

[0085] (3) Maintenance plan adjustment constraints

[0086] For any conventional hydro-thermal power unit i, it is necessary to determine whether it maintains the initial maintenance plan, that is:

[0087]

[0088] (4) Fan maintenance capacity balance equation

[0089] The actual maintenance capacity of the wind farm is determined by the maintenance 0-1 variable and the predicted maintenance capacity, which systematically takes into account the joint effects of the model optimization process and the external multi-attribute meteorological factors.

[0090] β w (t) = β′ w (t) b w (t) (17)

[0091] In the formula, β′ w (t) is the predicted maintenance capacity of wind farm w at time t, which is obtained from historical maintenance data and system operation status and is a known condition of the model; b w (t) is the wind farm maintenance 0-1 variable; b w (t) = 1 means that the wind farm w is under maintenance at time t, b w (t)=0 means that the wind farm w is in a non-maintenance state at time t.

[0092] (5) Consistency constraints on wind turbine maintenance capacity

[0093] According to the assumption 3 in step 101, within the optimized maintenance interval, the maintenance capacity should remain fixed and be the minimum value of the interval:

[0094]

[0095] Where t′ sta,w , t′ end,w are the start time and end time of the actual maintenance plan of wind farm w, respectively.

[0096] (6) Continuous constraints on fan maintenance

[0097]

[0098] (7) Total duration constraints for fan maintenance

[0099]

[0100] Where, T cx,w is the maintenance period of wind farm w. According to assumption 2, the present invention takes T cx,w =3. f is the number of times the wind power needs to be repaired within the repair planning period, and in this embodiment, f=1.

[0101] (8) Minimum interval constraints for fan maintenance

[0102] If a wind farm needs to undergo multiple maintenances within the maintenance planning cycle, according to the principles of "periodic maintenance" and "fault evolution law" in the actual maintenance system, it should be avoided that the time periods of two adjacent maintenances are too close. To this end, it is necessary to set the minimum time interval between two adjacent maintenances. The maintenance minimum interval constraint is set as follows:

[0103]

[0104] Where, T D is the minimum interval between two adjacent maintenances; τ is the time index, τ∈[1,T].

[0105] 3. System operation control constraints

[0106] (1) Upper and lower limits of thermal power generation unit power

[0107] (1-a i (t))P i,min ≤P i (t)≤(1-a i (t))P i,max (twenty two)

[0108] Where P i,min , P i,max(i∈Ω Th ) are the minimum technical output and maximum technical output of thermal power unit i respectively.

[0109] (2) Upper and lower limits of hydropower unit power

[0110] 0≤P i (t)≤(1-a i (t))P i,max (twenty three)

[0111] Where P i,max (i∈Ω Tm ) is the maximum technical output of hydropower unit i.

[0112] (3) Wind farm power constraints

[0113] 0≤P w,t ≤P′ w,t ·(1-γ t b w (t)) (24)

[0114] Where P w,t , P′ w,t They are the dispatched output and predicted output of wind farm w respectively.

[0115] (4) Power balance constraints

[0116]

[0117] Where P h,ex (t) is the interconnection power between the power grid in this region and the external power grid during period t; H is the total number of interconnection lines; D(t) is the daily maximum load in this region, which is the total load value of the system.

[0118] (5) System backup constraints

[0119]

[0120] In the formula, θ is the optimal standby coefficient set according to the system operation characteristics, and the present invention takes θ=0.05.

[0121] (6) System power flow constraints

[0122]

[0123] Where P L,l，max is the maximum transmission power of transmission line l; L is the total number of transmission lines; P 1,i,l (t) is the product of the power transfer factor between unit i and line l and the output of unit i, P 2,h,l (t) is the product of the power transfer factor between tie line h and line l and the power of tie line h, P 3,d,l(t) is the product of the power transfer factor between load node d and line l and the load value of node d, P 4,w,l (t) is the product of the power transfer factor between wind turbine w and line l and the output of wind turbine w.

[0124] (7) Contractual power constraints

[0125] When the annual maintenance planning cycle is used, the medium- and long-term trading cycles that need to be considered are annual and monthly, so the contractual electricity constraints include annual and monthly deviation electricity constraints.

[0126] For annual contracts, the deviation between the contracted electricity and the total market transaction electricity should be controlled within the acceptable range for both parties to the transaction, that is, the annual deviation electricity constraint is:

[0127]

[0128] Where D y is the number of months in a year, D y =12; d is the natural month variable; is the contract electricity deviation coefficient, which is 5%; E(t) is the electricity provided by the seller every day; E′(d) is the total electricity traded per month.

[0129] The annual contracted electricity is decomposed into monthly quantities, and the monthly electricity deviation is limited to a range acceptable to both parties. That is, the monthly deviation electricity constraint is:

[0130]

[0131] Where T′ is the total number of days in a natural month.

[0132] In order to reduce the redundancy of the symmetry interval described in the conventional IGDT model and eliminate the subjectivity of decision-making, the present invention introduces the optimal chance constraint theory considering probability, and by setting multiple target confidence levels, corrects the symmetry interval of uncertain variables, and achieves the preset target with relevant probability. The solution results under different confidence levels are compared and analyzed, and the optimal maintenance decision plan is selected according to the criterion of minimizing the objective function.

[0133] The uncertainty of wind power output depends on the changing characteristics of wind speed with climate, geographical location, etc. The study of wind power output uncertainty can be regarded as the study of wind speed uncertainty. In general, the probability distribution function of wind speed can be described as a two-parameter Weil distribution, and its probability density function and cumulative probability distribution are as follows:

[0134]

[0135]

[0136] In the formula, k w,t 、cw,t They are the shape and scale parameters of the Will distribution of wind speed at time t, which are fitted from historical data.

[0137] Since the mathematical relationship between wind power output and wind speed at the location is positively correlated, the probability distribution of wind speed can be substituted into the mathematical relationship to obtain the probability distribution of wind power output. The mathematical relationship between wind power output and wind speed is:

[0138]

[0139] In the formula, v w,r 、v w,o 、v w,N are the cut-in wind speed, cut-out wind speed and rated wind speed of the wind turbines in wind farm w respectively; P w,max is the rated output power of the wind turbine in wind farm w.

[0140] For the probability density function of wind speed, given the confidence level τ, its probability confidence interval Figure 2 Shown

[0141] Under a given confidence level, the preset deviation coefficient of wind power output is:

[0142]

[0143]

[0144]

[0145] In the formula, α w,t is the preset deviation coefficient of the output of wind farm w at time t; They are respectively the wind speed range and wind power output range under a given confidence level τ; They are the upper and lower bounds of the confidence interval of wind power output, respectively.

[0146] The fluctuation range of wind power output is:

[0147] U(α w,t ,P w,t ′)=[(1-α w,t ) w,t ′,(1+α w,t ) w,t ′] (36)

[0148] After the above processing, the preset deviation coefficient α under the confidence level is used w,t Instead of global optimization The wind power output range is determined by the envelope constraint model to be solved Transformed into U(α) under a given confidence level w,t ,P w,t ′)=[(1-αw,t ) w,t ′,(1+α w,t ) w,t ′], further refining the output range and achieving targeted uncertainty analysis under the condition of sufficient information.

[0149] Wind power output has a time-varying probability distribution, and its output range fluctuates over time. The same fixed value in all time periods, the CC-IGDT model does not take into account the deviation coefficient The time-varying characteristics of the system may cause the following problems when trying to maximize the avoidance of uncertainty risks in all time periods:

[0150] (1) Limitations of the optimization process. Conventional IGDT has a unique As the maximum deviation in each period is optimized, the entire optimization process only avoids the maximum possible deviation in a certain period, and the actual deviation in other periods is used to formulate the corresponding optimization strategy in the worst scenario according to this optimization direction. It can be regarded as the prediction curve y′(y′=f(P w,t ′)) is within the predicted output deviation boundary range [y d ,y u ](depending on the specific optimization problem) moves up and down and approaches the result of the deviation boundary. The curve movement process does not take into account the imbalance and irregular movement within the time period, and the optimization process is relatively limited.

[0151]

[0152] (2) The decision result is “conservative”. Although the uncertainty global deviation coefficient of the CC-IGDT model The risk can be avoided to the maximum extent, but at a higher confidence level, the volatility range of wind power output is expanded. At this time, there is an optimal deviation coefficient of wind power output in certain periods of time. Deviation coefficient from the global The difference is big, that is:

[0153]

[0154] Due to the characteristics of the CC-IGDT model, the time period t is forced to be optimized in a direction that is not conducive to economic efficiency, and the actual optimization deviation of each time period is close to the deviation boundary. In the robust model, the total system cost under more severe deviation scenarios is often higher. That is, the CC-IGDT model still has a high degree of system decision conservatism. This reduces the flexible adjustment function of the system and causes additional operating cost losses due to the superposition of uncertainties, which is not in line with the economic efficiency of the system operation.

[0155] In view of this, the present invention constructs a time-varying CC-IGDT model: considering the deviation coefficient The time-varying characteristics of the global deviation coefficient Decoupling time periods In order to maintain the equivalence of time scale with the optimization goal, the decoupled deviation coefficient Perform mean normalization to obtain the optimal deviation coefficient considering time-varying characteristics In order to achieve the goal of "accurately finding the optimal area and reducing the conservatism of decision-making". At the same time, through the duality theorem and uncertainty operation rules, the time-varying CC-IGDT model is converted into an equivalent deterministic one, specifically:

[0156]

[0157] In the formula, To consider the expected deviation coefficient of time-varying characteristics; is the time-varying deviation coefficient (changes with time); Ψ w,t -1 (τ) is the inverse cumulative distribution function of wind power output; p t is the deviation coefficient for period t The weight of; T is the maintenance planning period (T = 366); μ w is the deviation weight factor of wind farm w.

[0158] The principle of time-varying CC-IGDT to accurately find the optimal area and reduce the conservatism of decision-making is as follows:

[0159] (1) Optimization process based on information support (improving IGDT)

[0160] Probability distribution prediction of the output boundary of time-varying CC-IGDT (taking the above boundary as an example) u The geometric relationship between the envelope surface y1 considering the time-varying characteristics of the output (time-varying CC-IGDT) and the envelope surface y2 not considering the time-varying characteristics of the output (IGDT) is as follows: Figure 3 As shown, the mathematical expression is as follows:

[0161] y u =f(α w,t ,P w,t ′) (40)

[0162]

[0163]

[0164] In the formula, Considering the time-varying characteristics, relative to the preset deviation coefficient α w,t The offset factor set; κ is the preset deviation coefficient α without considering the time-varying characteristics w,t ’s offset factor; X1 and X2 are the independent variable sets of y1 and y2 respectively.

[0165] from Figure 3 It can be seen that the independent variable set of the envelope surface y1 is a subset of the independent variable set of the envelope surface y2, and the envelope curve y1 is based on the optimization direction of each time period. u Perform complex decoupling and translation operations, while y2 only refers to the optimization characteristics of a certain period of time. u Perform simple translation operations. Based on the formation principle of the deviation envelope surface, the time-varying CC-IGDT realizes the optimization process based on information support and accurately defines the optimization area.

[0166] The CC-IGDT model has a high degree of system decision conservatism because: the optimal deviation coefficient in certain periods of time Forced to set a worse deviation coefficient This results in additional operating costs.

[0167] Comparative analysis of the feasible domain of maximizing risk avoidance of uncertain variables in the time-varying CC-IGDT and CC-IGDT models:

[0168]

[0169] For the feasible region Υ1, the integral mean value theorem is used to obtain:

[0170]

[0171] Because the whole time Therefore Υ1≤Υ2. When Υ1<Υ2, the feasible domain of the time-varying CC-IGDT is smaller ( Figure 4 ), the model tends to be adjusted in a more economical direction, and the cost of solving is lower, that is,

[0172] From the above, we can see that the time-varying CC-IGDT can avoid the additional operating cost caused by CC-IGDT by considering the time-varying characteristics of the optimal deviation coefficient. The cost function expression is:

[0173]

[0174] Additional costs avoided for any period of time Taylor expansion of it:

[0175]

[0176] As the total time period (planning period T) increases, C loss It presents complex high-dimensional nonlinear characteristics, and the additional cost avoided by time-varying CC-IGDT is higher, which reduces the conservatism of decision-making to a certain extent.

[0177] In summary, compared with the prior art, the present invention has the following technical advantages:

[0178] 1) The combined low-carbon maintenance optimization model of cascade hydropower-wind-thermal units proposed for the first time in this invention takes the operation of the power grid as the perspective, fully considers the coupling constraints of cascade hydropower, system operation constraints, maintenance constraints, contract power constraints, etc., realizes the staggered maintenance of hydropower and thermal power, seasonal maintenance of wind power, and combined maintenance of wind power and cascade hydropower and thermal power, making up for the deficiency of the single optimization object of the maintenance model described in the existing literature.

[0179] 2) The optimization goal of the joint low-carbon maintenance optimization model of cascade hydropower-wind-thermal units proposed in the present invention takes into account the low-carbon demand for maintenance optimization. The solved maintenance plan helps to accommodate wind power and can effectively realize the joint low-carbon maintenance optimization of multiple types of units under large power grids.

[0180] 3) The time-varying CC-IGDT method proposed for the first time in this invention is innovative and realizes the optimization process based on information support and the optimal opportunity constraint based on probability, providing a new theoretical basis and research ideas for uncertainty optimization.

[0181] 4) The combined low-carbon maintenance optimization model of cascade hydropower-wind-thermal units based on time-varying CC-IGDT constructed for the first time in this invention fully considers the impact of wind power output uncertainty on unit maintenance, and can obtain a maintenance plan that adapts to uncertain disturbances.

[0182] 5) The time-varying CC-IGDT method described in the present invention can always obtain a maintenance plan with a cost lower than the expected cost, that is, the time-varying CC-IGDT can still seek a more economical maintenance plan after taking into account the uncertainty of wind power output, and the maintenance optimization process fully considers the priority wind power consumption characteristics and system operation characteristics.

[0183] 6) Compared with the conventional IGDT method, the time-varying CC-IGDT method of the present invention realizes an optimization process based on information support, which sacrifices the partial robustness of the maintenance plan and the deviation coefficient to the decision cost to obtain a larger arbitrage space, and the required decision cost is lower. Compared with the scenario method, the maintenance plan optimized by the time-varying CC-IGDT method has stronger adaptability to random scenarios and shows better computational efficiency.

[0184] 7) The present invention proposes a combined maintenance optimization method for cascade hydropower-wind-thermal units taking into account uncertainty. Compared with the existing maintenance optimization methods, the present invention solves the problem of combined maintenance of fans and hydropower units, and provides decision makers with environmentally adaptable maintenance plans under different risk preferences.

[0185] The above embodiments are only for illustrating the technical concept and features of the present invention, and their purpose is to enable ordinary technicians in the field to understand the content of the present invention and implement it accordingly, and they cannot be used to limit the protection scope of the present invention. Any equivalent changes or modifications made based on the essence of the content of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for optimizing the combined maintenance of cascade hydropower-wind-thermal units taking into account uncertainty, characterized in that: include: Construct a low-carbon maintenance optimization model for cascaded hydropower-wind-thermal units, and solve the model to obtain the basic maintenance plan and basic cost; Based on the wind speed data and combined with the historical wind speed data at historical moments, the probability distribution function of wind speed is solved; The probability distribution function of wind power output is obtained according to the probability distribution function of wind speed, and the wind power output range and minimum deviation coefficient under different confidence levels are solved; Based on the basic cost and the minimum deviation coefficient, and taking the wind power output range under different confidence levels as the uncertainty range, a cascade hydropower-wind-thermal unit joint low-carbon maintenance optimization model based on time-varying CC-IGDT is constructed, including: Global deviation coefficient Decoupling time periods The time-varying deviation coefficient after decoupling Normalize the mean to obtain the expected deviation coefficient considering the time-varying characteristics Through the duality theorem and uncertainty operation rules, the time-varying CC-IGDT model is converted into an equivalent deterministic one; The combined low-carbon maintenance optimization model of the cascade hydropower-wind-thermal units based on time-varying CC-IGDT is solved to obtain the total cost and the combined low-carbon maintenance optimization plan of the cascade hydropower-wind-thermal units.

2. The method for optimizing the combined maintenance of cascade hydropower-wind-thermal units taking into account uncertainty according to claim 1, characterized in that: The cascade hydropower-wind-thermal unit joint low-carbon maintenance optimization model is constructed in the following way: Ignoring the uncertainty of wind power, with the goal of minimizing economic losses and facilitating low-carbon operation of the power system, the unit maintenance cost, operating cost, and carbon emission cost targets are comprehensively considered, and the optimization target is constructed by the weighted sum of various costs. Combined with the constraints, a joint low-carbon maintenance optimization model for cascaded hydropower-wind-thermal units is constructed.

3. The combined maintenance optimization method for cascade hydropower-wind-thermal units taking into account uncertainty as claimed in claim 1, characterized in that: The total cost is: Unit maintenance cost It is expressed as: Unit operating cost It is expressed as: Carbon emissions costs It is expressed as: Where, T is the maintenance period; i is the unit number; Z is the total number of hydropower units, Z = Th + Tm, Th, Tm are the number of thermal power units and hydropower units respectively; W is the total number of wind farms; C jx,i,t , C jx,w,t Respectively represent the maintenance costs of unit capacity water-fired power units and fans; P i,max is the maximum output of the hydro-thermal power unit; a i (t) is the maintenance 0-1 variable of unit i; a i (t) = 1 means that unit i is under maintenance at time t, a i (t) = 0 means that unit i is in a non-maintenance state at time t; β w (t) is the maintenance continuous variable of the wind farm w, and its value reflects the capacity value that actually needs maintenance in the wind farm; P i (t) is the output of thermal power unit i at time t; c 0,i 、c 1,i 、c 2,i is the coal consumption coefficient of thermal power units; is the carbon cost per kWh, δ t is the time interval; are the weights of unit maintenance cost, operating cost and carbon emission cost respectively.

4. The combined maintenance optimization method for cascade hydropower-wind-thermal units taking into account uncertainty as claimed in claim 2, characterized in that: The constraints include cascade hydraulic and electric power coupling constraints, unit maintenance constraints, and system operation control constraints.

5. The combined maintenance optimization method for cascade hydropower-wind-thermal units taking into account uncertainty as claimed in claim 4, characterized in that: The cascade hydroelectric coupling constraints include upper and lower limit constraints on hydropower unit output, upper and lower limit constraints on hydropower station reservoir capacity, upper and lower limit constraints on hydropower station power generation flow, abandoned water volume constraints, and ecological flow constraints.

6. The combined maintenance optimization method for cascade hydropower-wind-thermal units taking into account uncertainty as claimed in claim 4, characterized in that: The unit maintenance constraints include maintenance continuity constraints, maintenance duration constraints, maintenance plan adjustment constraints, fan maintenance capacity balance equations, fan maintenance capacity consistency constraints, fan maintenance continuity constraints, fan maintenance total duration constraints, and fan maintenance minimum interval constraints.

7. The method for optimizing the combined maintenance of cascade hydropower-wind-thermal units taking into account uncertainty as claimed in claim 4, characterized in that: The system operation control constraints include upper and lower power constraints of thermal power units, upper and lower power constraints of hydropower units, wind farm power constraints, power balance constraints, system reserve constraints, system flow constraints, and contract power constraints.

8. The method for optimizing the combined maintenance of cascade hydropower-wind-thermal units taking into account uncertainty according to claim 1, characterized in that: The probability distribution function of the wind speed is described as a two-parameter Weil distribution, and its probability density function and cumulative probability distribution are as follows: In the formula, k w,t 、c w,t They are the shape and scale parameters of the Will distribution of wind speed at time t, which are fitted from historical data.

9. The combined maintenance optimization method for cascade hydropower-wind-thermal units taking into account uncertainty as claimed in claim 8, characterized in that: The method of obtaining the probability distribution function of wind power output according to the probability distribution function of wind speed and solving the wind power output interval and the minimum deviation coefficient under different confidence levels includes: Substituting the probability distribution function of wind speed into the mathematical relationship, we can get the probability distribution function of wind power output. The mathematical relationship between wind power output and wind speed is: In the formula, v w,r 、v w,o 、v w,N are the cut-in wind speed, cut-out wind speed and rated wind speed of the wind turbines in wind farm w respectively; P w,max is the rated output power of the wind turbine in wind farm w; For the probability density function of wind speed, under a given confidence level τ, the preset deviation coefficient of wind power output is: In the formula, α w,t is the preset deviation coefficient of the output of wind farm w at time t; They are respectively the wind speed range and wind power output range under a given confidence level τ; They are the upper and lower bounds of the confidence interval of wind power output respectively; The fluctuation range of wind power output is: U(a w,t ,P w,t ′)=[(1-a w,t )P w,t ′,(1+α w,t )P w,t ′] Use the preset deviation coefficient α under confidence level w,t Instead of global optimization The wind power output range is determined by the envelope constraint model to be solved Transformed into U(α) under a given confidence level w,t ,P w,t ′)=[(1-α w,t ) w,t ′,(1+α w,t ) w,t ′].

10. The combined maintenance optimization method of cascade hydropower-wind-thermal units taking into account uncertainty as claimed in claim 8, characterized in that: The time-varying CC-IGDT model is converted into an equivalent deterministic model through the duality theorem and uncertainty operation rules, specifically: In the formula, To consider the expected deviation coefficient of time-varying characteristics; is the time-varying deviation coefficient; w,t -1 (τ) is the inverse cumulative distribution function of wind power output; p t is the deviation coefficient for period t The weight of; T is the maintenance planning cycle; μ w is the deviation weight factor of wind farm w.

Images