A method and system for optimizing unit combination

Abstract

This invention discloses a unit combination optimization method and system. Based on the output characteristics analysis of different types of generating units, a mathematical model of each unit in the power system is constructed. The baseline state of each unit is obtained from the mathematical model. Building upon existing unit combination research, this method balances the goals of renewable energy consumption and optimal economic cost, comprehensively considering the unit itself, the system, and constraints. It improves computational efficiency while enhancing the matching degree between the unit combination plan and the grid transmission capacity, reserving space for economic dispatch and safety verification. A heuristic algorithm with low computational complexity and ease of operation provides a feasible approach to the above objectives. By arranging the start-up, shutdown, and output states of the units, the method effectively avoids the solution process of mixed integer programming problems, improving the computational efficiency of long-cycle unit combination problems in large-scale practical power systems, reducing computation time from several hours to a few minutes.

Description

Technical Field

[0001] This invention belongs to the field of unit combination scheduling technology, specifically relating to a unit combination optimization method and system. Background Technology

[0002] Unit combination is a crucial component of power system dispatch optimization, aiming to minimize total cost by determining the start-up and shutdown status and output plans of each unit within the dispatch cycle. The construction of a new generation of power systems with a high proportion of renewable energy integration is a vital step in achieving my country's energy transition. However, the integration of renewable energy also significantly impacts the safe and economical operation of the power system, posing greater challenges to unit combination calculations. It is necessary to consider the goal of maximizing the absorption of uncertain renewable energy sources and the impact of cybersecurity constraints on operational planning, while also taking into account the computational performance of long-term evaluation and optimization. Therefore, it is essential to research an efficient solution method for the unit combination optimization problem of multi-generational power generation that comprehensively considers various constraints.

[0003] Extensive research has been conducted on methods for solving unit combination problems, primarily focusing on heuristic algorithms, mathematical optimization algorithms, and intelligent optimization algorithms. The first two categories are classic algorithms for solving unit combination problems and have been widely applied. Heuristic algorithms rely on intuitive judgment or practical scheduling experience, considering local optima at each time granularity to assume the final solution is globally optimal. Typical heuristic algorithms include exhaustive search and priority ordering. Mathematical optimization algorithms mainly include Lagrange relaxation and mixed integer programming. The former involves relaxing system constraints by adding penalty terms to the objective function, then decomposing the unit combination problem into a series of single-unit subproblems and dual problems. The single-unit subproblems are solved using dynamic programming. Even problems are solved using the subgradient method; the latter is a class of mathematical methods for solving problems with both integer and non-integer variables. Representative methods include the Benders decomposition method, which decomposes the problem into two subproblems containing only discrete variables and only continuous variables, and iterates between these two subproblems using a coordinating factor to finally obtain the optimal solution; and the branch and bound method, which implicitly enumerates all possible combinations and can theoretically obtain the optimal solution. Intelligent optimization algorithms achieve optimization by simulating known evolutionary methods in nature, such as genetic algorithms, simulated annealing algorithms, and particle swarm optimization algorithms.

[0004] The aforementioned algorithms can demonstrate significant advantages under specific system and operational requirements, but they still possess certain technical limitations and shortcomings. Intelligent optimization is essentially stochastic optimization; when dealing with large-scale real-world systems, the increased model dimensionality significantly increases computational complexity, leading to slow calculations and insufficient stability of the obtained solutions. Under the same conditions, two solutions may yield different results, affecting the fairness of scheduling and the interpretation of optimization results. The Lagrange multiplier method suffers from poor flexibility in adding constraints and difficulty in converging a large number of Lagrange multipliers during iteration. Due to the complexity of constraints, it is difficult to comprehensively and meticulously consider all constraints of the system network and the generating units themselves, potentially resulting in unrealizable power generation plans. With the development of commercial solution software, optimization methods have gradually become a research hotspot; however, they remain limited by computational speed in long-term simulations of large power systems. If optimization methods are used in long-term simulations on an annual scale, a large number of variables are referenced along with numerous constraints, and the computationally intensive nature of mixed-integer programming makes it difficult to provide timely references for operators. Summary of the Invention

[0005] The purpose of this invention is to provide a unit combination optimization method and system to overcome the problems of slow speed and low efficiency in solving unit combination optimization problems using mixed integer programming.

[0006] A unit combination optimization method includes the following steps:

[0007] S1. Based on the output characteristics analysis of different types of generating units, construct mathematical models of each generating unit in the power system, and obtain the baseline state of each generating unit based on the mathematical models of each generating unit.

[0008] S2. Compare the available spinning reserve capacity under the baseline state with the minimum reserve capacity required by the constraints. Check whether the units in the system can simultaneously meet the power balance constraints and spinning reserve capacity constraints when each unit is in start-up and output state without considering ramp-up constraints. If not, start-up operation is performed until the power balance constraints and spinning reserve capacity constraints are met simultaneously, and the corrected baseline state is obtained.

[0009] S3. Determine the remaining load required for each operating unit under the corrected baseline state, calculate the upward ramp capacity and downward ramp capacity of each operating unit, compare the relationship between the ramp capacity and the remaining load of each operating unit, and if the operating units under the current corrected baseline state cannot meet the remaining load when considering ramp constraints, then start or shut down the units under the principle of optimal economy until the operating units under the current corrected baseline state meet the remaining load when considering ramp constraints, and then allocate the remaining load to each operating unit.

[0010] Preferably, the baseline status of each unit includes the start-up / shutdown status, output status, and the current output limit of the new energy units in each unit.

[0011] Preferably, the initial state is taken as the reference state for each unit, and the state at time t-1 is taken as the reference state for each unit at time t-1.

[0012] Preferably, the initial state is determined with reference to the optimization solution result at the median load size within one week; the upper limit of the output of the new energy unit at any time is the predicted output of the new energy unit at that time.

[0013] Preferably, the available spinning reserve capacity under the reference state is the remaining spinning reserve capacity after the unit output meets the power balance constraint under the reference state. If the available spinning reserve capacity under the reference state is less than the lower limit requirement of spinning reserve, additional units need to be started until the reserve capacity constraint is met, and the output of the additional units is limited to its minimum technical output value, and finally the corrected reference state is obtained.

[0014] Preferably, the expression for calculating the remaining load is as follows:

[0015]

[0016] In the formula, α represents the renewable energy unit absorption rate. To adjust the output of thermal power units from the baseline state, To contribute to the timing of new energy generating units Let t be the original load at time t.

[0017] Preferably, the unit's ramp-up capacity is zero when the unit is stopped; when the unit is started, the difference between the baseline output and the output limit is calculated, and the unit's ramp-up capacity is the maximum of the above difference and the hourly ramp-up capacity.

[0018] Preferably, if the remaining load is greater than zero, the relationship between the remaining load and the upward ramp capacity is compared; if the remaining load is less than or equal to zero, it is compared with the downward ramp capacity. When the upward ramp capacity or the downward ramp capacity cannot meet the remaining load, the corresponding start-up or shutdown operation needs to be performed.

[0019] Preferably, when the upward ramp capacity can meet the remaining load, the unit start-up and shutdown status is not changed; when the downward ramp capacity can meet the remaining load, the spinning reserve capacity is checked to see if it exceeds the upper limit of the spinning reserve. If the spinning reserve capacity is too large, the unit needs to be shut down.

[0020] A unit combination optimization system includes a preprocessing module and an optimization module.

[0021] The preprocessing module analyzes the output characteristics of different types of generating units, constructs mathematical models of each generating unit in the power system, and obtains the baseline state of each generating unit based on the mathematical models of each unit.

[0022] The optimization module compares the available spinning reserve capacity under the baseline state with the minimum reserve capacity required by the constraints. It examines whether the units in the system, without considering ramp-up constraints, can simultaneously meet the power balance and spinning reserve capacity constraints under the start-up, shutdown, and output states of each unit. If not, it performs start-up operations until both constraints are met, thus obtaining the corrected baseline state. It then determines the remaining load required by each unit under the corrected baseline state, calculates the upward and downward ramp-up capacities of each unit, and compares the ramp-up capacities with the remaining load. If, under the current corrected baseline state, the units cannot meet the remaining load under ramp-up constraints, it performs start-up or shutdown operations based on the principle of optimal economy until the units meet the remaining load under ramp-up constraints. Finally, it allocates the remaining load to each unit.

[0023] Compared with the prior art, the present invention has the following beneficial technical effects:

[0024] This invention presents a unit combination optimization method. Based on the analysis of the output characteristics of different types of units, a mathematical model of each unit in the power system is constructed. The baseline state of each unit is obtained based on the mathematical model. Building upon existing unit combination research, this method balances the goals of renewable energy consumption and optimal economic cost. It comprehensively considers the unit itself, the system, and constraints, improving computational efficiency while enhancing the matching degree between the unit combination plan and the grid transmission capacity, reserving space for economic dispatch and safety verification. A heuristic algorithm with low computational complexity and ease of operation provides a feasible approach to the above objectives. By arranging the start-up, shutdown, and output states of units, the method effectively avoids the solution process of mixed integer programming problems, improving the computational efficiency of long-cycle unit combination problems in large-scale practical power systems, reducing computation time from several hours to a few minutes.

[0025] This invention achieves the goal of maximizing the absorption of new energy sources and optimizing costs by arranging the output of different types of generating units and setting the switching sequence of thermal power units.

[0026] By considering multiple levels of constraints, taking into account the details of unit operation and grid DC power flow constraints, the final power generation plan can be effectively matched with the actual system grid transmission capacity.

[0027] This invention, by considering the minimum start-up and shutdown time of the units and the minimum technical output limit of each unit when formulating strategies, can avoid the predicament of some economically inefficient units being unable to start for a long time when using optimization methods. The strategy formulation method is more in line with the actual decision-making of dispatchers and can fully schedule all planned units. Attached Figure Description

[0028] Figure 1 This is a schematic diagram of the unit combination optimization method in an embodiment of the present invention.

[0029] Figure 2 This is the process for rotating spare capacity under the verification baseline state in this embodiment of the invention.

[0030] Figure 3 This is the process for comparing the remaining load with the unit's ramp-up capacity in an embodiment of the present invention.

[0031] Figure 4 This is a flowchart of the heuristic unit combination strategy algorithm in an embodiment of the present invention.

[0032] Figure 5 This is a test case network topology diagram in an embodiment of the present invention.

[0033] Figure 6 (a) Figure 6 (b) are line graphs showing the predicted output of wind power and photovoltaic units in the embodiments of the present invention. Detailed Implementation

[0034] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0035] like Figure 1 As shown, the present invention provides a unit combination optimization method, comprising the following steps:

[0036] S1. Based on the output characteristics analysis of different types of generating units, construct mathematical models of each generating unit in the power system, and obtain the baseline state of each generating unit based on the mathematical models of each generating unit.

[0037] The baseline status of each unit includes the start-up and shutdown status, output status, and the current output limit of the new energy units in each unit;

[0038] S2, Verify the spinning reserve capacity under the baseline state: Compare the available spinning reserve capacity under the baseline state with the minimum reserve capacity required by the constraints. Check whether the units in the system can simultaneously meet the power balance constraints and spinning reserve capacity constraints when the ramp-up constraints are not considered, under the start-up and output states of each unit. If not, start-up operation is performed until the power balance constraints and spinning reserve capacity constraints are simultaneously met, and the corrected baseline state is obtained.

[0039] S3, Verify the ramp capacity under the corrected baseline state; determine the remaining load required by each operating unit under the corrected baseline state, calculate the upward and downward ramp capacity of each operating unit, compare the relationship between the ramp capacity of each operating unit and the remaining load, if each operating unit under the current corrected baseline state cannot meet the remaining load when considering ramp constraints, then perform start-up or shutdown operations under the principle of optimal economy until each operating unit under the current corrected baseline state meets the remaining load when considering ramp constraints, and then allocate the remaining load to each operating unit.

[0040] This invention determines a baseline state based on the current time; verifies whether the available spinning reserve capacity under the baseline state can meet the lower limit constraint of the spinning reserve; if not, it performs a startup operation to obtain a corrected baseline state; next, it determines the remaining load that the unit needs to bear under the corrected baseline state and the ramp-up capacity of the computer group; finally, it compares the ramp-up capacity of the started-up units with the remaining load; if the current start-up / shutdown state of the units cannot meet the remaining load under the ramp-up constraint, it performs a startup or shutdown operation until each started-up unit under the current corrected baseline state meets the remaining load when considering the ramp-up constraint, and then distributes the remaining load to each started-up unit.

[0041] In step S1, the initial state is taken as the baseline state for each unit, and the state at time t-1 is taken as the baseline state for each unit. When formulating the start-up and shutdown plan for each unit at the initial time, the minimum start-up and shutdown time and ramp-up constraints do not need to be considered. The determination of the initial state is based on the median load size within one week, with the optimization solution result as the reference benchmark. The upper limit of the output of the new energy units at any time is the predicted output of the new energy units at that time.

[0042] The purpose of solving the load median time optimization problem within a week is to incorporate constraints and make the results more realistic. Using a weekly unit, the time when the weekly load is the median is selected. The objective function is the lowest economic cost after piecewise linearization, and a mixed-integer programming problem is solved for that specific time. Using the obtained unit start-up and shutdown states and output states as reference benchmarks, the unit combination for each time point within the week is arranged sequentially forward. This method provides a relatively effective reference for formulating weekly unit combination plans with the fewest optimization calculations.

[0043] Piecewise linearization of the objective function will also improve computational efficiency. For example, the coal consumption function can be divided into m segments, and the objective function is as follows:

[0044]

[0045] Among them, K i,s C represents the slope of each segment of the piecewise linearized coal consumption function. 0,i p represents the coal consumption generated when the unit is started and operating at minimum output. i,t,s The output of the unit in stages must satisfy the following constraints:

[0046]

[0047] When performing long-cycle unit combination calculations with a time scale greater than one week, the initial time needs to consider whether the changes in the unit state at the minimum load time of the current week and the unit state at the last time of the previous week meet the unit's own technical constraints. Since the hourly ramp-up rate of thermal power units is relatively fast, only thermal power units whose start-up and shutdown states have changed can be checked to see if they can meet the minimum continuous start-up and shutdown time constraints.

[0048] In step S2, considering the coupling relationship between system-level power balance and reserve capacity constraints, the available spinning reserve capacity under the baseline state is assumed to be the remaining spinning reserve capacity after the unit output meets the power balance constraint under the baseline state, i.e., the output of all thermal power units at time t. Power output from new energy units The sum equals the load size at the current moment. When the thermal power unit is started, the available standby capacity is... Equations (3) and (4) below represent the spinning reserve capacity constraint and power balance constraint, respectively. Substituting expression (4) into (3) yields the available spinning reserve capacity under the baseline condition, as shown in equation (5):

[0049]

[0050]

[0051]

[0052] The procedure for verifying the available spare capacity under the baseline condition is as follows: Figure 2 As shown, if the available capacity of the spinning reserve under the baseline state is less than the lower limit requirement of the spinning reserve, additional units need to be started until the reserve capacity constraint is met, and the output of the additional units is limited to its minimum technical output value, thus obtaining the corrected baseline state.

[0053] Specifically, when adding generating units, the most economical unit should be started first, in order of economic efficiency.

[0054] Step S3, the verification of the climbing capacity under the baseline condition, specifically includes the following steps:

[0055] Calculate the magnitude of the remaining load: The magnitude of the remaining load determines the target of adjusting the start-up and shutdown status and output status of the thermal power unit, that is, the output gap that needs to be filled.

[0056] Original load at time t It is divided into three parts, one part being the output of the thermal power unit under the modified reference state. The remaining portion is satisfied by the output of new energy units; to simplify the calculation, the time-series output of new energy units can be expressed as the predicted output value multiplied by an empirical proportionality coefficient, i.e. The last part is the unmet load, i.e., the remaining load P. left,t The expression for calculating the remaining load is as follows:

[0057]

[0058] In the formula, α is the renewable energy unit absorption rate, which is set to 90% in this application.

[0059] Calculating the ramp-up capacity of thermal power units: Factors affecting the ramp-up capacity of a unit include the current start-up / shutdown status of the unit, the upper limit of output, the lower limit of output, and the ramp-up rate. In the shutdown state, the ramp-up capacity is zero. In the start-up state, the difference between the baseline output and the output limit is calculated. The ramp-up capacity of the unit is the maximum or minimum of the above difference and the hourly ramp-up capacity, as shown in the following formula:

[0060] Uramp i =min{P i,max ,(P i,t-1 +R u )} (7)

[0061] Dramp i =max{P i,min ,(P i,t-1 -R d )} (8)

[0062] In the formula, Uramp i Dramp i q represents the upward ramp capacity of thermal power unit i and the downward ramp capacity of thermal power unit q, respectively.

[0063] Comparison of remaining load and unit ramp-up capacity: The comparison process is as follows (see attached). Figure 3 As shown, if the remaining load is greater than zero, the relationship between the remaining load and the upward ramp capacity is compared; if the remaining load is less than or equal to zero, it is compared with the downward ramp capacity.

[0064] When the ramp-up or ramp-down capacity cannot meet the remaining load, corresponding start-up or shutdown operations are required. When the ramp-up capacity can meet the remaining load, the unit start-up / shutdown status remains unchanged. When the ramp-down capacity can meet the remaining load, it is checked whether the spinning reserve capacity exceeds the upper limit of the spinning reserve. If the spinning reserve capacity is too large, a shutdown operation is required. The ramp-up process takes into account improving the absorption of renewable energy. When ramping up, priority is given to increasing the output of renewable energy units to the upper limit. When ramping down, if the capacity is sufficient, the output of renewable energy units is increased to replace part of the output of thermal power units. Simultaneously, the load shortfall caused by shutdown is compensated primarily by increasing the output of renewable energy units.

[0065] Start-up or shutdown operations: To avoid reducing computational efficiency due to repetitive looping processes, the first step in starting up or shutting down is to identify operable units based on their current status and operating conditions. Then, the units are started up or shut down in a predetermined sequence until system constraints are met. Operable units are determined based on their own technical constraints, such as units whose baseline state for start-up is shutdown and whose continuous shutdown time meets the minimum continuous downtime constraint, and whose minimum technical output is less than the current remaining load. The order of operations is determined by minimizing economic costs. Units with better economic efficiency are prioritized for start-up, while units are shut down in reverse economic order for shutdown. For ease of calculation and rational arrangement, this application adopts an approximate method of prioritizing the operation of larger-capacity units and prioritizing the shutdown of smaller-capacity units, assuming that the economic efficiency of a thermal power unit is directly proportional to its capacity.

[0066] When the ramp-up capacity of the unit is sufficient to meet the remaining load, the remaining load is allocated to each operating unit. The unit combination process described in this patent aims to determine the start-up and shutdown status of the thermal power unit. The fine allocation and arrangement of unit output will be carried out during the economic dispatch process. In this step, the output of each unit is determined by an approximate method of allocating the load proportionally based on the ramp-up space of each operating unit.

[0067] Example

[0068] The following is a flowchart of the heuristic unit combination strategy algorithm (attached). Figure 4 The steps of this invention are described in detail below:

[0069] Determine unit start-up, shutdown, and output: Based on the input data, merge the unit maintenance plan, mark the units under maintenance at each time, calculate the predicted output value of the new energy units at each time, and set it as their output upper limit. Obtain the unit baseline state at time t, i.e., the unit start-up and shutdown state, continuous start-up and shutdown time, and output at time t-1, and correct the upper limit of the new energy unit output to the predicted value at time t.

[0070] The verification process checks whether the available spinning reserve capacity under the baseline state meets the lower limit constraint of spinning reserve. Specifically, under the baseline unit start-up and shutdown conditions, assuming the unit output meets the power balance constraint, and the output power is equal to the current load, whether the remaining spinning reserve capacity of the started units meets the lower limit requirement. If not, a start-up operation is performed. The thermal power units that meet the minimum continuous downtime constraint among the currently not started units are selected and started up in descending order of capacity, operating at the minimum technical output value, until the lower limit requirement of spinning reserve is met, thus obtaining the corrected baseline state.

[0071] Calculate the remaining load that the unit needs to bear under the corrected baseline state, that is, the remaining unmet load at the current moment after deducting the two parts of the original load that are satisfied by the output of thermal power units under the baseline state and the load that is consumed by the output of new energy units according to empirical proportions.

[0072] Compare the relationship between the remaining load and the ramp-up capacity, and consider four scenarios to determine whether to perform unit start-up or shutdown operations. If the remaining load is greater than zero, compare the remaining load with the ramp-up capacity. If the ramp-up capacity can meet the remaining load, do not change the unit's start-up / shutdown status. If the ramp-up capacity cannot meet the remaining load, start-up operation is required. Select the thermal power units that meet the minimum continuous downtime constraint among the currently not started units, and whose minimum technical output value is less than the remaining load. Start up the units in descending order of capacity and operate at the minimum technical output value. Adjust the remaining load once for each unit put into operation until the remaining load is less than zero. If the remaining load is less than zero, compare the absolute value of the remaining load with the size of the downward ramp capacity. When the downward ramp capacity can meet the remaining load, check whether the spinning reserve capacity exceeds the upper limit of the spinning reserve. If the spinning reserve capacity is too large, a shutdown operation is required. Select the thermal power units that meet the minimum continuous downtime constraint among the currently running units and shut them down in order of increasing capacity until the upper limit of the spinning reserve capacity is met. The remaining load needs to be adjusted once for each unit shut down. When the downward ramp capacity cannot meet the remaining load, a shutdown operation is also required. The remaining load needs to be adjusted once for each unit shut down until the remaining load is greater than zero.

[0073] The remaining load obtained in the above four scenarios is allocated as follows: For the remaining load greater than zero that can be satisfied by ramping up, priority is given to increasing the output of new energy sources to the upper limit, and the remaining load is allocated proportionally to the generating units for ramping up; for the remaining load greater than zero that cannot be satisfied by ramping up, the corrected remaining load less than zero obtained after starting the unit is allocated proportionally to the generating units for ramping down; for the remaining load less than zero that can be satisfied by ramping down, the corrected remaining load obtained after adjusting for the upper limit constraint of spinning reserve is allocated proportionally to the generating units for ramping down; for the remaining load less than zero that cannot be satisfied by ramping down, the corrected remaining load greater than zero obtained after shutting down the unit is allocated proportionally to the generating units for ramping up.

[0074] Record the start-up, shutdown, and power output of the units at time t, update the continuous start-up and shutdown time of each thermal power unit, and then proceed to the next time step. It is assumed that the optimal value at each time step constitutes the global optimal value. Finally, check whether the loop termination condition is met.

[0075] To illustrate the applicability and effectiveness of the heuristic strategy unit combination optimization method provided by this invention, specific examples are provided below:

[0076] This invention is simulated in an improved IEEE 24-node test system, with the network topology as follows: Figure 5 As shown, four wind power plants and four photovoltaic power plants were added. Two concentrated solar power plants and two pumped storage power plants were also added to the system. The simulation time step was set to 1 hour, and a week-long unit combination optimization calculation was performed. The predicted wind power and photovoltaic power generation data were adjusted based on the actual output of the wind power and photovoltaic power plants. The predicted weekly power generation data for wind and photovoltaic units can be found in [the following data is missing from the original text]. Figure 6 The results obtained in (a) and (b) are shown in Tables 1 and 2. The computation time of the proposed heuristic strategy and the solution time and economic cost of solving the optimization problem using the CPLEX solver under the same environment are also presented.

[0077] Table 1 Heuristic Strategy and Optimization Solution Computation Schedule

[0078] Optimize solution time / s Heuristic strategy solution time / s speed ratio 41.943 0.718 58.42

[0079] Table 2. Heuristic Strategies and Economic Costs of Optimization Solutions

[0080] Optimized economic cost / yuan Heuristic economic cost / yuan deviation / % <![CDATA[4.5764×10 4 ]]> <![CDATA[4.7377×10 4 ]]> 3.25

[0081] Therefore, using heuristic strategies for unit combination optimization calculations can significantly improve efficiency and control the result error within a certain range, reducing the calculation time from several hours to a few minutes, and providing timely reference for operators.

Claims

1. A method for optimizing unit combination, characterized in that, Includes the following steps: S1. Based on the output characteristics analysis of different types of generating units, construct mathematical models of each generating unit in the power system, and obtain the baseline state of each generating unit based on the mathematical models of each generating unit. S2. Compare the available spinning reserve capacity under the baseline state with the minimum reserve capacity required by the constraints. Check whether the units in the system can simultaneously meet the power balance constraints and spinning reserve capacity constraints when each unit is in start-up and output state without considering ramp-up constraints. If not, start-up operation is performed until the power balance constraints and spinning reserve capacity constraints are met simultaneously, and the corrected baseline state is obtained. S3, determine the remaining load that each unit needs to bear under the corrected reference state, calculate the upward ramp capacity and downward ramp capacity of each unit, compare the relationship between the ramp capacity of each unit and the remaining load, if each unit cannot meet the remaining load under the current corrected reference state when considering ramp constraints, then start or shut down the unit under the principle of optimal economy until each unit meets the remaining load under the current corrected reference state when considering ramp constraints, and then allocate the remaining load to each unit. If the remaining load is greater than zero, compare the remaining load with the uphill capacity; if the remaining load is less than or equal to zero, compare it with the downhill capacity. When the ramp-up or ramp-down capacity cannot meet the remaining load, perform the corresponding start-up or shutdown operation.

2. The unit combination optimization method according to claim 1, characterized in that, The baseline status of each unit includes the start-up and shutdown status, output status, and the current output limit of the new energy units in each unit.

3. The unit combination optimization method according to claim 1, characterized in that, Initially, the reference state is used as the baseline state for each unit. t At all times t The state at time -1 is the baseline state for each unit.

4. The unit combination optimization method according to claim 3, characterized in that, The initial state is determined with reference to the median load size within a week and the optimization solution results; the upper limit of the output of the new energy units at any time is the predicted output of the new energy units at that time.

5. The unit combination optimization method according to claim 2, characterized in that, The available spinning reserve capacity under the baseline state is the remaining spinning reserve capacity after the unit output meets the power balance constraint under the baseline state. If the available spinning reserve capacity under the baseline state is less than the lower limit requirement of the spinning reserve, additional units are started until the reserve capacity constraint is met, and the output of the additional units is limited to its minimum technical output value, and finally the corrected baseline state is obtained.

6. A unit combination optimization method according to claim 2, characterized in that, The expression for calculating the remaining load is as follows: (6) In the formula, For the absorption rate of new energy units, To adjust the output of thermal power units from the baseline state, To contribute to the timing of new energy generating units for t Original load at any given moment.

7. The unit combination optimization method according to claim 2, characterized in that, When the unit is stopped, its ramp-up capacity is zero; when the unit is running, the difference between the baseline output and the output limit is calculated, and the ramp-up capacity of the unit is the maximum value of the above difference and the hourly ramp-up capacity.

8. The unit combination optimization method according to claim 1, characterized in that, When the upward ramp capacity can meet the remaining load, the unit start-up and shutdown status is not changed; when the downward ramp capacity can meet the remaining load, the spinning reserve capacity is checked to see if it exceeds the upper limit of the spinning reserve. If the spinning reserve capacity is too large, the unit needs to be shut down.

9. A unit combination optimization system, characterized in that, Includes a preprocessing module and an optimization module. The preprocessing module analyzes the output characteristics of different types of generating units, constructs mathematical models of each generating unit in the power system, and obtains the baseline state of each generating unit based on the mathematical models of each unit. The optimization module compares the available spinning reserve capacity under the baseline state with the minimum reserve capacity required by the constraints. It examines whether the units in the system, without considering ramp-up constraints, can simultaneously meet power balance and spinning reserve capacity constraints under the start-up, shutdown, and output states of each unit. If not, it performs start-up operations until both constraints are met, thus obtaining the corrected baseline state. It then determines the remaining load required by each unit under the corrected baseline state, calculates the upward and downward ramp-up capacities of each unit, and compares the ramp-up capacities with the remaining load. If, under the current corrected baseline state, the units cannot meet the remaining load under ramp-up constraints, it performs start-up or shutdown operations based on the principle of optimal economy until the units meet the remaining load under ramp-up constraints. The remaining load is then allocated to each unit. If the remaining load is greater than zero, it compares the remaining load with the upward ramp-up capacity; if the remaining load is less than or equal to zero, it compares it with the downward ramp-up capacity. When the ramp-up or ramp-down capacity cannot meet the remaining load, perform the corresponding start-up or shutdown operation.

Images