Unit commitment heuristic solving method based on priority method

Abstract

The invention belongs to the technical field of unit commitment solving, and particularly relates to a unit commitment heuristic solving method based on a priority method, which comprises the following steps: constructing an objective function of a power system unit commitment model, and setting constraint conditions; sorting the units by using the average full-load cost as a sorting factor; obtaining an initial solution of a unit combination problem; utilizing a unit adjustment strategy based on the minimum start-stop time constraint to adjust unit output and start-stop states; setting a simple redundant unit adjustment strategy to adjust unit output and start-stop states to obtain an approximate solution of a unit combination problem; adjusting the output and start-stop states of the unit by using a unit adjustment strategy based on the climbing constraint; setting an accurate redundant unit adjustment strategy, and adjusting unit output and start-stop states; and performing output adjustment between the units to obtain a final solution, and calculating the minimum operation cost. According to the method, the solving efficiency of the power system unit combination can be improved, and the quality of the unit combination solution can be improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of unit commitment solution, and particularly relates to a heuristic solution method for unit commitment based on the priority order method. Background Art

[0002] Unit commitment in a power system refers to formulating the start-stop and power generation plans of units (i.e., generating units) with the goal of minimizing cost (consumption) within a certain scheduling period (usually one day or one week), achieving system power balance and meeting certain constraint conditions and reserve requirements. Mathematically, the unit commitment problem belongs to a large-scale mixed integer programming problem, and it is difficult to obtain an exact optimal solution within an effective time.

[0003] Currently, the methods commonly used to solve the unit commitment problem mainly include heuristic algorithms, mathematical optimization algorithms, and intelligent optimization algorithms. The results obtained by mathematical optimization methods are relatively accurate, but the amount of calculation in the solution process is large and the solution time is long. Intelligent optimization algorithms can learn and predict based on previous data, improving the solution speed, but they are prone to falling into local optima during the decision-making process. Heuristic algorithms such as the priority order method and the reverse sorting method mainly start from intuitive judgment and practical experience, and use some heuristic strategies to solve the unit commitment problem within a limited search space. Among them, the priority order method determines the sorting index according to the economic characteristics of each generating unit, and arranges the unit switching and load distribution in sequence according to the sorting index, with a relatively fast convergence speed. However, the solution obtained by the priority order method based on considering the power balance constraint and the spinning reserve constraint often cannot meet the requirements of the unit output boundary constraint, the minimum start-stop time constraint, and the ramping constraint at the same time, resulting in a low solution quality. To improve the solution quality, the priority order method often adopts a heuristic adjustment strategy to handle the unit output boundary constraint and the minimum start-stop time constraint, and often combines with a mathematical optimization algorithm or an intelligent optimization algorithm when dealing with the ramping constraint, making it difficult to avoid the disadvantages of other algorithms.

[0004] Therefore, it is necessary to design a heuristic solution method for unit commitment based on the priority order method that can handle the ramping constraint to improve the solution efficiency of unit commitment in a power system. Summary of the Invention

[0005] According to the above deficiencies in the prior art, the purpose of the present invention is to provide a heuristic solution method for unit commitment based on the priority order method, which can improve the solution efficiency of unit commitment in a power system and can improve the quality of the unit commitment solution.

[0006] To achieve the above object, the present invention provides a heuristic solution method for unit commitment based on the priority order method, including the following steps: S1. With the goal of minimizing fuel costs and startup costs, construct the objective function of the unit commitment model for the power system, and set the unit output boundary constraints, system power balance constraints, spinning reserve constraints, unit ramp rate constraints, startup and shutdown logic constraints, and minimum startup and shutdown time constraints; S2. Use the average full-load cost as the sorting factor and sort the units in ascending order; S3. Determine the startup and shutdown status of the units to meet the spinning reserve constraints, and then determine the unit output to meet the system power balance constraints and unit output boundary constraints, thereby obtaining the initial solution to the unit commitment problem; S4. Use the unit adjustment strategy based on the minimum startup and shutdown time constraints to adjust the unit output and startup and shutdown status; S5. Set the simple redundant unit adjustment strategy to adjust the unit output and startup and shutdown status to obtain an approximate solution to the unit commitment problem; S6. Use the unit adjustment strategy based on the ramp rate constraints to adjust the unit output and startup and shutdown status; S7. Set the precise redundant unit adjustment strategy to adjust the unit output and startup and shutdown status; S8. Perform output adjustment between units to obtain the final solution and calculate the minimum operating cost.

[0007] As a preferred solution of the present invention, in the above S1, the objective function OF of the unit commitment model for the power system is: ; In the formula, N T is the total optimization period, and t represents one of the moments; N G is the number of units, and g represents one of the units; is the output of unit g at time t; , are 0-1 integer variables, representing the startup and shutdown status of unit g at time t and time t-1 respectively. The value of 1 means the unit is on, and the value of 0 means the unit is off; is the startup cost of unit g; is the fuel cost of unit g; The expression of is: ; In the formula, , , are the power generation cost coefficients of unit g; The expression of is: ; In the formula, , They are the hot start-up cost and cold start-up cost of unit g respectively; is the continuous shutdown time of unit g at time t; is the minimum shutdown time of unit g; is the cold start-up time of unit g.

[0008] As a preferred solution of the present invention, in the above S1, the constraint conditions are specifically: Output boundary constraint of the unit: ; In the formula, and represent the minimum and maximum values of the output of unit g respectively; System power balance constraint: ; In the formula, represents the load at time t; Spinning reserve constraint: ; In the formula, represents the reserve capacity at time t; Unit ramp rate constraint: ; ; In the formula, is the output of unit g at time t - 1; and represent the power increase and decrease rate limits of unit g respectively; and represent the power-up rate limit and power-down rate limit of unit g respectively; and are both 0 - 1 integer variables, represents the start-up operation of unit g at time t, with a value of 1 indicating that the start-up operation is performed at time t and a value of 0 indicating that the start-up operation is not performed; represents the shutdown operation of unit g at time t, with a value of 1 indicating that the shutdown operation is performed at time t and a value of 0 indicating that the shutdown operation is not performed; Start-stop logic constraint: ; ; Minimum start-stop time constraint, including minimum start-up time constraint and minimum shutdown time constraint: ; ; In the formula, and Indicates a 0-1 integer variable, representing the start-up and shutdown operations of unit g at time s respectively; Is the minimum start-up time of unit g; Represents the universal quantifier.

[0009] As a preferred embodiment of the present invention, in S2, the average full-load cost refers to the average power generation cost of the unit in the full-load state, and numerically is the power generation cost of the unit in the full-load state divided by the maximum output of the unit. For unit g, its average full-load cost Is expressed as: .

[0010] As a preferred embodiment of the present invention, in S3, the process of obtaining the initial solution of the unit commitment is as follows: S3.1. Determine the start-stop state of the unit to meet the spinning reserve constraint. Assume that the start-stop state of all units at each time is 0. The process is as follows: S3.1.1. Let t = 1; S3.1.2. Let g = 1, representing the first unit after sorting; S3.1.3. Let , Is the sum of the maximum outputs of the units started at time t; S3.1.4. Let ; S3.1.5. If , then , let g = g + 1. If at this time , go to step S3.1.4, otherwise go to step S3.1.6, where Represents the last unit after sorting; If , then , go to step S3.1.6; S3.1.6. Let t = t + 1. If , go to step S3.1.2, otherwise end the adjustment; S3.2. Determine the unit output to meet the system power balance constraint and the unit output boundary constraint, so as to obtain the initial solution of the unit commitment problem. Assume that the output of all units at each time is 0. The process is as follows: S3.2.1. Let t = 1; S3.2.2. Let g = 1; S3.2.3. Let , Is the sum of the powers of the units started at time t; S3.2.4. If , go to step S3.2.6, otherwise , at this time, if , then , if , then ; S3.2.5. Let g = g + 1. If , go to step S3.2.4; otherwise, go to step S3.2.6; S3.2.6. Let t = t + 1. If , go to step S3.2.2; otherwise, end the adjustment.

[0011] As a preferred solution of the present invention, in S4, the output and start - stop state of the unit are adjusted by using the unit adjustment strategy based on the minimum start - stop time constraint. Specifically, if the minimum start - up time constraint cannot be satisfied at a certain moment, the unit is started at the next moment; if the minimum shutdown time constraint cannot be satisfied at a certain moment, the unit is started at that moment; when the state of unit g changes from 0 to 1 at time t, the output of unit g is .

[0012] As a preferred solution of the present invention, in S5, a simple redundant unit adjustment strategy is set to adjust the output and start - stop state of the unit to obtain an approximate solution to the unit commitment problem. Specifically, when the sum of the unit outputs exceeds the total load demand, the system has output redundancy, and at this time, the output of the redundant unit needs to be adjusted; during the adjustment process, the simple redundant unit adjustment strategy considers the influence of the unit output boundary constraint, the system power balance constraint, and the spinning reserve constraint, and does not involve the unit ramp - rate constraint; starting from the unit with the lowest priority, the process is as follows: S5.1.1. Let t = 1; S5.1.2. Calculate the redundant output at time t : ; S5.1.3. Let ; S5.1.4. Let , , represents the adjustable amount of unit g, is the total adjustment amount at time t; S5.1.5. Calculate : ; S5.1.6. If , then , ; Let g = g - 1. If g ≥ 1, go to step S5.1.5; otherwise, go to step S5.1.7; If , then , go to step S5.1.7; S5.1.7. Let t = t + 1. If , go to step S5.1.2; otherwise, end the adjustment.

[0013] As a preferred solution of the present invention, in S6, the output and start-stop state of the unit are adjusted by using a unit adjustment strategy based on ramp constraints. Specifically, the time period and the unit are classified and processed according to the start-stop change of the unit; the continuous on-period of the unit is recorded as a "1" type time period, and the continuous off-period of the unit is recorded as a "0" type time period; the unit that remains on throughout the time period is called a constant-on unit, and the unit that remains off throughout the time period is called a constant-off unit, and the unit with start-stop changes is called a start-stop unit; among them, the constant-off unit and the constant-on unit that maintains the maximum output constantly throughout the time period always meet the unit ramp constraints, so no adjustment is required; For the constant-on unit, adjust the output of the constant-on unit with a changing output throughout the time period to meet the unit ramp constraints. Since this type of unit is a "1" type time period throughout the time period, the method of "increasing in the forward direction and checking in the reverse direction" is adopted to adjust the output of the unit to meet the unit ramp constraints. The specific process is as follows: S6.1.1. Increase in the forward direction, and the steps include: S6.1.1.1. Let , where is the initial moment of the "1" type time period; S6.1.1.2. If , then ; If , then ; S6.1.1.3. Let t = t + 1. If , go to step S6.1.1.2; otherwise, end the adjustment, where is the end moment of the "1" type time period; S6.1.2. Check in the reverse direction, and the steps include: S6.1.2.1. Let ; S6.1.2.2. If , then ; S6.1.2.3. Let t = t - 1. If , go to step S6.1.2.2; otherwise, end the adjustment; For start-stop units, adjust the output and start-stop status of the start-stop units to meet the unit ramp constraints; the start-stop units include both "1" type time periods and "0" type time periods throughout the entire period; first, adopt the method of "increasing in the forward direction and verifying in the reverse direction" for the "1" type time periods to adjust the unit output to meet the unit ramp constraints; then, adjust the "0" type time periods of the start-stop units at different times. The specific process is as follows: S6.2.1. For the start time, when the "0" type time period contains the start time, it is necessary to adjust the output and start-stop status of the unit at the end time of the "0" type time period; assume that the "0" type time period is "1 to t 1 ", t 1 is the end time of the "0" type time period, then: S6.2.1.1. Let t = t 1 ; S6.2.1.2. If , then and , and go to step S6.2.1.3, otherwise end the adjustment, where is the output of unit g at time t + 1; S6.2.1.3. Let t = t - 1. If t ≥ 1, go to step S6.2.1.2, otherwise end the adjustment; S6.2.2. For the end time, when the "0" type time period contains the end time, it is necessary to adjust the output and start-stop status of the unit at the start time of the "0" type time period; assume that the "0" type time period is "t 2 to N T ", t 2 is the start time of the "0" type time period, then: S6.2.2.1. Let t = t 2 ; S6.2.2.2. If , then and , and go to step S6.2.2.3, otherwise end the adjustment, where is the output of unit g at time t - 1; S6.2.2.3. Let t = t + 1. If t ≤ N T , go to step S6.2.2.2, otherwise end the adjustment; S6.2.3. For the intermediate time, when the "0" type time period only contains intermediate times, it is necessary to adjust the output and start-stop status of the units at both the start and end times of the "0" type time period; assume that the "0" type time period is "t 3 to t 4 ", t 3 and t 4 are the start and end times of the "0" type time period respectively, then: S6.2.3.1. For the left - end adjustment of the "0" - type time period, let \(t = t\) 3 ; S6.2.3.2. If , then and , go to step S6.2.3.3; otherwise, end the left - end adjustment; S6.2.3.3. Let \(t=t + 1\). If \(t\leq t\) 4 , go to step S6.2.3.2; otherwise, end the left - end adjustment; S6.2.3.4. For the right - end adjustment of the "0" - type time period, let \(t = t\) 4 ; S6.2.3.5. If , then and , go to step S6.2.3.6; otherwise, end the right - end adjustment; S6.2.3.6. Let \(t=t - 1\). If \(t\geq t\) 3 , go to step S6.2.3.5; otherwise, end the right - end adjustment; S6.2.3.7. After the adjustments at both ends, calculate the shutdown duration of the units in the "0" - type time period at this time ; if , it means that the "0" - type time period still meets the minimum shutdown time constraint, and the adjustment ends; if , it means that the "0" - type time period cannot meet the minimum shutdown time constraint; to maintain the system power balance, combine this "0" - type time period with the adjacent "1" - type time periods before and after into a "1" - type time period; finally, adopt the method of "increasing in the forward direction and checking in the reverse direction" to adjust the unit output of the newly formed "1" - type time period to meet the unit ramp - up constraint.

[0014] As a preferred solution of the present invention, in S7, using the precise redundant unit adjustment strategy, on the basis of meeting each constraint condition, adjust the unit output and start - stop state at each moment in turn to obtain the final solution of the unit commitment problem. The specific process is as follows: Since the unit output at the head and tail moments is affected by the unit ramp - up constraint of the adjacent moment at one end, and the intermediate moments are affected by the unit ramp - up constraints of the adjacent moments at both ends, the adjustment process divides the entire time period into a head period including the head moment, an intermediate period including only intermediate moments, and a tail period including the tail moment for redundant unit adjustment; For the head period: S7.1.1. Let \(t = 1\); S7.1.2. Calculate the redundant output \(R\) of the head moment 1 : ; In the formula, Indicates the output of unit g at the head-end moment; Indicates the load at the head-end moment; S7.1.3. Let ; S7.1.4. Let and ; S7.1.5. Calculate the adjustable amount of unit g: ; In the formula, Indicates the output of unit g at the right-end moment of the head-end period; S7.1.6. If , then and , let g = g - 1. If g ≥ 1, go to step S7.1.5; otherwise, end the adjustment; If , then , end the adjustment; For the middle period: S7.2.1. Let t = 2; S7.2.2. Calculate the redundant output at time t: ; S7.2.3. Let ; S7.2.4. Let and ; S7.2.5. Calculate the adjustable amount of unit g: ; S7.2.6. If , then and , let g = g - 1. If g ≥ 1, go to step S7.2.5; otherwise, go to step S7.2.7; If , then , go to step S7.2.7; S7.2.7. Let t = t + 1. If , then go to step S7.2.2; otherwise, end the adjustment; For the end period, the adjustment process for the end period is the same as that for the head-end period. Only the influence of the output of one end unit needs to be considered, that is, consider the output of the unit at the left-end moment of the end period.

[0015] As a preferred embodiment of the present invention, in S8, after the adjustment of redundant units in the head period, the middle period, and the end period is completed, the sum of the unit outputs at each moment is fixed. At this time, it is necessary to adjust the output between units. The adjustment method is to divide the whole period into the head period, the middle period, and the end period, and adjust by increasing the output of priority units and decreasing the output of non-priority units; among them, the adjustment process in the middle period is as follows: S8.1. Let t = 2; S8.2. Let g = 1; S8.3. If , go to step S8.9, otherwise: ; In the formula, represents the power that the priority unit can increase; S8.4. Let , , represents the power that the non-priority unit can decrease; represents the total reduction of the non-priority unit power; S8.5. Let , where i is the index of the unit; S8.6. If , then: ; Otherwise, go to step S8.8; Among them, is a 0-1 integer variable representing the start-stop state of unit i at time t; , , represent the output of unit i at time t, time t-1, and time t+1 respectively; , represent the power increase and decrease speed limits of unit i respectively; is the minimum output of unit i; S8.7. If , then: ; ; ; Go to step S8.8; If , then: ; ; Go to step S8.10; S8.8. Let \(i = i - 1\). If \(i>g\), go to step S8.6; otherwise, go to step S8.9; S8.9. Let \(g = g + 1\). If , go to step S8.3; otherwise, go to step S8.10; S8.10. Let \(t = t + 1\). If , go to step S8.2; otherwise, end the adjustment; For the first - end time period and the last - end time period, since they are only affected by the unit ramp - up / down constraints of the adjacent time - moments at one end, the single - end adjustment can be carried out in the same way as the adjustment process of the middle time period. After the adjustment is completed, calculate the minimum operating cost of the unit commitment problem according to the obtained unit output and start - stop status.

[0016] The algorithm involved in the present invention can be executed by an electronic device. The electronic device includes a memory, a processor, and a computer program stored on the memory and executable on the processor. The above - mentioned algorithm calculation is realized by the processor executing the software.

[0017] The beneficial effects of the present invention are as follows: By introducing heuristic adjustment strategies, such as the minimum start - stop time constraint adjustment strategy (adjustment of the unit adjustment strategy based on the minimum start - stop time constraint), the simple redundant unit adjustment strategy, and the ramp - up / down constraint adjustment strategy (unit adjustment strategy based on the ramp - up / down constraint), the present invention effectively solves the problem that the traditional priority - order method may violate the constraint conditions in the initial solution. These strategies can gradually optimize the output and start - stop status of the units, making the final solution meet the strict requirements in multiple aspects such as system power balance, spinning reserve, unit output boundary, minimum start - stop time, and unit ramp - up / down constraint, thus significantly improving the quality of the unit commitment solution.

[0018] While ensuring a relatively high calculation accuracy, the present invention greatly improves the solution efficiency. Through the staged optimization adjustment, the complex iterative calculation process is avoided, and a solution close to the optimal solution can be obtained in a short time. For example, in the simulation verification of a 10 - unit test system, compared with the commercial solver Gurobi, although the operating cost is slightly higher by 0.59%, the solution time is reduced by nearly 40.7%. This shows that the method is particularly suitable for solving the unit commitment problem with high time requirements and has strong practicability and economy. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 is the flow schematic diagram of the present invention; Figure 2 is the schematic diagram of the output curves of 10 generating units in the initial solution during the verification process of the present invention; Figure 3It is a schematic diagram of the start-stop states of 10 generating units in the approximate solution during the verification process of the present invention; Figure 4 It is a schematic diagram of the output curves of 10 generating units in the approximate solution during the verification process of the present invention; Figure 5 It is a schematic diagram of the output curves of 10 generating units in the final solution during the verification process of the present invention. Specific embodiments

[0020] The following further describes the embodiments of the present invention with reference to the accompanying drawings: As Figure 1 shown, a heuristic solution method for unit commitment based on the priority order method includes the following steps: S1. With the goal of minimizing fuel cost and start-up cost, construct the objective function of the power system unit commitment model, and set the unit output boundary constraints, system power balance constraints, spinning reserve constraints, unit ramp rate constraints, start-stop logic constraints, and minimum start-stop time constraints; S2. Use the average full-load cost as the sorting factor and sort the units in ascending order; S3. Determine the start-stop states of the units to meet the spinning reserve constraints, and then determine the unit outputs to meet the system power balance constraints and unit output boundary constraints, thereby obtaining the initial solution to the unit commitment problem; S4. Use the unit adjustment strategy based on the minimum start-stop time constraint to adjust the unit outputs and start-stop states; S5. Set a simple redundant unit adjustment strategy to adjust the unit outputs and start-stop states to obtain an approximate solution to the unit commitment problem; S6. Use the unit adjustment strategy based on the ramp rate constraint to adjust the unit outputs and start-stop states; S7. Set an accurate redundant unit adjustment strategy to adjust the unit outputs and start-stop states; S8. Perform output adjustment between units to obtain the final solution and calculate the minimum operating cost.

[0021] In S1, the objective function OF of the power system unit commitment model is: ; In the formula, N T is the total optimization period, and t represents one of the moments; N G is the number of units, and g represents one of the units; is the output of unit g at time t; , are 0-1 integer variables, representing the start-stop states of unit g at time t and time t-1 respectively. A value of 1 indicates start-up, and a value of 0 indicates shutdown; is the start-up cost of unit g; is the fuel cost of unit g; The expression of is: In the formula, , , are the power generation cost coefficients of unit g, which are well-known contents. Specifically: is the quadratic term coefficient, which represents the non-linear growth part of the fuel cost when the unit output increases, reflects the efficiency change of the unit during high-load operation, and is usually used to describe the additional fuel consumption of the unit under high load; is the linear term coefficient, which represents the linear growth part of the fuel cost when the unit output increases, reflects the unit output cost of the unit, and is directly related to the fuel consumption rate; is the constant term, which represents the fixed cost of the unit during operation. Even when the output of the unit is zero, this part of the cost will still be generated. This part of the cost is usually related to the maintenance and start-up costs of the unit; The expression of is: In the formula, , are the hot start-up cost and cold start-up cost of unit g respectively; is the continuous shutdown time of unit g at time t; is the minimum shutdown time of unit g; is the cold start-up time of unit g.

[0022] In this embodiment, for time and time period, time only includes one time, such as the 1st time and the 5th time. A time period is usually composed of continuous times. For example, the 1st to 2nd times are a time period, and the 5th to 9th times are a time period. Setting time periods is to facilitate the analysis of the changes in unit output and start-stop at different times within each time period. For example, adjusting the unit output and start-stop status at a certain time in a certain time period. For the total optimization time period, N T takes the value of 24, and the total optimization time period includes the 1st time to the 24th time, representing one day.

[0023] In S1, the constraint conditions are specifically: Unit output boundary constraint: ; In the formula, , represent the minimum value and maximum value of the output of unit g respectively; System power balance constraint: ; In the formula, represents the load at time t; Spinning reserve constraint: ; In the formula, represents the reserve capacity at time t; Unit ramp rate constraint: ; ; In the formula, is the output of unit g at time t - 1; , respectively represent the power increase and decrease speed limits of unit g; and respectively represent the start-up power speed limit and shutdown power speed limit of unit g; and are both 0 - 1 integer variables, represents the start-up operation of unit g at time t, with a value of 1 indicating that the start-up operation is executed at time t and a value of 0 indicating that the start-up operation is not executed; represents the shutdown operation of unit g at time t, with a value of 1 indicating that the shutdown operation is executed at time t and a value of 0 indicating that the shutdown operation is not executed; Start-stop logic constraint: ; ; Minimum start-stop time constraint, including minimum start-up time constraint and minimum shutdown time constraint: ; ; In the formula, , represent 0 - 1 integer variables, respectively representing the start-up and shutdown operations of unit g at time s (similarly to , ); is the minimum start-up time of unit g; represents the universal quantifier.

[0024] In S2, the average full-load cost refers to the average power generation cost of the unit in the full-load state. Numerically, it is the power generation cost of the unit in the full-load state divided by the maximum output of the unit. For unit g, its average full-load cost is expressed as: .

[0025] In S3, the process of obtaining the initial solution of the unit commitment is as follows: S3.1. Determine the start-stop states of the units to meet the spinning reserve constraint. Assume that the start-stop states of all units at each moment are 0. The process is as follows: S3.1.1. Let t = 1; S3.1.2. Let g = 1, representing the first unit after sorting; S3.1.3. Let , is the maximum sum of the outputs of the units that are on at time t; S3.1.4. Let ; S3.1.5. If , then , let g = g + 1. If at this time , go to step S3.1.4, otherwise go to step S3.1.6, where represents the last unit after sorting; If , then , go to step S3.1.6; S3.1.6. Let t = t + 1. If , go to step S3.1.2, otherwise end the adjustment; S3.2. Determine the unit outputs to meet the system power balance constraint and the unit output boundary constraint, so as to obtain the initial solution of the unit commitment problem. Assume that the outputs of all units at each moment are 0. The process is as follows: S3.2.1. Let t = 1; S3.2.2. Let g = 1; S3.2.3. Let , is the sum of the powers of the units that are on at time t; S3.2.4. If , go to step S3.2.6, otherwise , at this time, if , then , if , then ; S3.2.5. Let g = g + 1. If , go to step S3.2.4, otherwise go to step S3.2.6; S3.2.6. Let t = t + 1. If , go to step S3.2.2, otherwise end the adjustment.

[0026] In S4, the output and start / stop status of the units are adjusted using the unit adjustment strategy based on the minimum start / stop time constraint. Specifically, if the minimum on-time constraint cannot be met at a certain moment, the unit is turned on at the next moment; if the minimum off-time constraint cannot be met at a certain moment, the unit is turned on at that moment; when the status of unit g changes from 0 to 1 at time t, the output of unit g is 。

[0027] In S5, a simple redundant unit adjustment strategy is set to adjust the output and start / stop status of the units, obtaining an approximate solution to the unit commitment problem. Specifically, when the sum of the unit outputs exceeds the total load demand, the system has output redundancy, and at this time, the output of the redundant units needs to be adjusted; during the adjustment process, the simple redundant unit adjustment strategy takes into account the influence of the unit output boundary constraint, the system power balance constraint, and the spinning reserve constraint, and does not involve the unit ramp rate constraint; in order to save operating costs, the adjustment starts from the unit with the lowest priority, and the process is as follows: S5.1.1. Let t = 1; S5.1.2. Calculate the redundant output at time t : ; S5.1.3. Let ; S5.1.4. Let , , represents the adjustable amount of unit g, is the total adjustment amount at time t; S5.1.5. Calculate : ; S5.1.6. If ,then , ;Let g = g - 1, if g ≥ 1, go to step S5.1.5, otherwise go to step S5.1.7; If ,then ,go to step S5.1.7; S5.1.7. Let t = t + 1, if ,go to step S5.1.2, otherwise end the adjustment.

[0028] In S6, the output and start / stop status of the units are adjusted using a unit adjustment strategy based on ramp constraints. Specifically, the time periods and units are classified separately according to the start / stop changes of the units. The continuous on-period of a unit is recorded as a "Type 1" time period, and the continuous off-period of a unit is recorded as a "Type 0" time period. Units that are on throughout the entire period are called constantly-on units, units that are off throughout the entire period are called constantly-off units, and units with start / stop changes are called start / stop units. Among them, constantly-off units and constantly-on units that maintain the maximum output constantly throughout the entire period always satisfy the unit ramp constraints, so no adjustment is required. For constantly-on units, adjust the output of constantly-on units with changing output throughout the entire period to satisfy the unit ramp constraints. Since such units are in "Type 1" time periods throughout the entire period, the method of "increasing in the forward direction and verifying in the reverse direction" is adopted to adjust the unit output to satisfy the unit ramp constraints. The specific process is as follows: S6.1.1. Increase in the forward direction, and the steps include: S6.1.1.1. Let , where is the initial moment of the "Type 1" time period; S6.1.1.2. If , then ; If , then ; S6.1.1.3. Let t = t + 1. If , go to step S6.1.1.2; otherwise, end the adjustment, where is the end moment of the "Type 1" time period; S6.1.2. Verify in the reverse direction, and the steps include: S6.1.2.1. Let ; S6.1.2.2. If , then ; S6.1.2.3. Let t = t - 1. If , go to step S6.1.2.2; otherwise, end the adjustment; For start / stop units, adjust the output and start / stop status of start / stop units to satisfy the unit ramp constraints. Start / stop units contain both "Type 1" time periods and "Type 0" time periods throughout the entire period. The "Type 0" time period may include the start moment (the 1st moment), the end moment (the 24th moment), or only intermediate moments (the 2nd - 23rd moments). First, adopt the method of "increasing in the forward direction and verifying in the reverse direction" for the "Type 1" time period to adjust the unit output to satisfy the unit ramp constraints. Then, adjust the "Type 0" time periods of start / stop units with different moments. The specific process is as follows: S6.2.1. For the first moment (the first moment refers to the 1st moment, the middle moment refers to any moment among the 2nd - 23rd moments, and the last moment refers to the 24th moment), when the "0" - type time period contains the first moment, it is necessary to adjust the output and start - stop status of the unit at the right - hand moment of the "0" - type time period; assume that the "0" - type time period is "1 to t 1 " (i.e., from the 1st moment to the t 1 th moment), t 1 is the right - hand moment of the "0" - type time period, then: S6.2.1.1. Let t = t 1 ; S6.2.1.2. If , then and , and go to step S6.2.1.3, otherwise end the adjustment, where is the output of unit g at the (t + 1)th moment; S6.2.1.3. Let t = t - 1, if t≥1, go to step S6.2.1.2, otherwise end the adjustment; S6.2.2. For the last moment (the 24th moment), when the "0" - type time period contains the last moment, it is necessary to adjust the output and start - stop status of the unit at the left - hand moment of the "0" - type time period; assume that the "0" - type time period is "t 2 to N T ", t 2 is the left - hand moment of the "0" - type time period, then: S6.2.2.1. Let t = t 2 ; S6.2.2.2. If , then and , and go to step S6.2.2.3, otherwise end the adjustment, where is the output of unit g at the (t - 1)th moment; S6.2.2.3. Let t = t + 1, if t≤N T , go to step S6.2.2.2, otherwise end the adjustment; S6.2.3. For the middle moment (the 2nd - 23rd moment), when the "0" - type time period only contains the middle moment, it is necessary to adjust the output and start - stop status of the units at both the left - hand and right - hand moments of the "0" - type time period; assume that the "0" - type time period is "t 3 to t 4 ", t 3 and t 4 are the left - hand and right - hand moments of the "0" - type time period respectively, then: S6.2.3.1. For the left - hand adjustment of the "0" - type time period, let t = t 3 ; S6.2.3.2. If , then and , go to step S6.2.3.3; otherwise, end the left - end adjustment. S6.2.3.3. Let \(t = t + 1\). If \(t\leq t\) 4 , go to step S6.2.3.2; otherwise, end the left - end adjustment. S6.2.3.4. For the right - end adjustment of the "0" - type time period, let \(t = t\) 4 ; S6.2.3.5. If , then and , go to step S6.2.3.6; otherwise, end the right - end adjustment. S6.2.3.6. Let \(t = t - 1\). If \(t\geq t\) 3 , go to step S6.2.3.5; otherwise, end the right - end adjustment. S6.2.3.7. After the adjustments at both ends, count the shutdown duration of the units in the "0" - type time period at this time ; if , it means that the "0" - type time period still meets the minimum shutdown time constraint, and the adjustment ends; if , it means that the "0" - type time period cannot meet the minimum shutdown time constraint; in order to maintain the system power balance, combine this "0" - type time period with the adjacent "1" - type time periods before and after into a "1" - type time period; finally, adopt the method of "increasing in the forward direction and checking in the reverse direction" to adjust the unit output of the newly formed "1" - type time period to meet the unit ramp - up constraint.

[0029] In S7, using the precise redundant unit adjustment strategy, on the basis of meeting each constraint condition, adjust the unit output and start - stop status at each moment in turn to obtain the final solution of the unit commitment problem. The specific process is as follows: Since the unit output at the head - end moment and the tail - end moment is affected by the unit ramp - up constraint of the adjacent moment at one end, and the intermediate moments are affected by the unit ramp - up constraints of the adjacent moments at both ends, the adjustment process divides the entire time period into a head - end period containing the head - end moment, an intermediate period containing only intermediate moments, and a tail - end period containing the tail - end moment for redundant unit adjustment; For the head - end period: S7.1.1. Let \(t = 1\); S7.1.2. Calculate the redundant output \(R\) of the head - end moment 1 : ; In the formula, represents the output of unit \(g\) at the head - end moment; represents the load at the head - end moment; S7.1.3. Let ; To ensure economy, reducing the output of redundant units should start from the least prioritized unit. S7.1.4. Let and ; S7.1.5. Calculate the adjustable capacity of unit g: ; In the formula, represents the right - hand end time of the first - end period (the first - end period only includes the 1st and 2nd moments, and at this time the right - hand end time is the 2nd moment, so it is represented as ), the output of unit g; S7.1.6. If , then and , let g = g - 1. If g≥1, go to step S7.1.5; otherwise, end the adjustment; If , then , end the adjustment; For the middle period: S7.2.1. Let t = 2; S7.2.2. Calculate the redundant output at time t: ; S7.2.3. Let ; S7.2.4. Let and ; S7.2.5. Calculate the adjustable capacity of unit g: ; S7.2.6. If , then and , let g = g - 1. If g≥1, go to step S7.2.5; otherwise, go to step S7.2.7; If , then , go to step S7.2.7; S7.2.7. Let t = t + 1. If , then go to step S7.2.2; otherwise, end the adjustment; For the last - end period, the adjustment process of the last - end period is the same as that of the first - end period. Only the influence of the output of one - end unit needs to be considered, that is, consider the output of the unit at the left - hand end time of the last - end period (the last - end period only includes the 23rd and 24th moments, and at this time the left - hand end time is the 23rd moment).

[0030] In S8, after the redundant unit adjustments in the initial period, intermediate period, and end period are completed, the sum of the unit outputs at each moment is fixed. To better reduce the unit operation cost, it is necessary to adjust the outputs among the units. The adjustment method is to divide the entire period into the initial period, intermediate period, and end period, and adjust by increasing the output of the priority units and reducing the output of the non-priority units. Among them, the adjustment process in the intermediate period is as follows: S8.1. Let t = 2; S8.2. Let g = 1; S8.3. If , go to step S8.9; otherwise: ; In the formula, represents the power that the priority unit can increase; S8.4. Let , , represents the power that the non-priority unit can reduce; represents the total reduction in the power of the non-priority units; The priority order method requires the priority units to output more power and the non-priority units to output less power; S8.5. Let , where i is the index of the unit (non-priority unit); Here, it means starting from the least priority unit (the last unit after sorting) to reduce the output of the non-priority units; S8.2 means starting from the most priority unit 1 to increase the output of the priority units, and the two finally reach a power balance; S8.6. If , then: ; Otherwise, go to step S8.8; Among them, is a 0-1 integer variable representing the start-stop state of unit i at time t ( is the same as ); , , , respectively represent the output of unit i at time t, t - 1, and t + 1; , respectively represent the power increase and decrease speed limits of unit i; is the minimum output of unit i; S8.7. If , then: ; ; Go to step S8.8; If , then: ; ; Go to step S8.10; S8.8. Let i = i - 1. If i > g, go to step S8.6; otherwise, go to step S8.9; S8.9. Let g = g + 1. If , go to step S8.3; otherwise, go to step S8.10; S8.10. Let t = t + 1. If , go to step S8.2; otherwise, end the adjustment; For the first - end time period and the last - end time period, the first - end time period and the last - end time period are only affected by the unit ramp - up / down constraints of the adjacent moment at one end. The single - end adjustment can be carried out in the same way as the adjustment process of the middle time period; After the adjustment is completed, calculate the minimum operating cost (the minimum of the objective function OF) of the unit commitment problem according to the obtained unit output and start - stop status. At this time, the final unit output and start - stop status can be obtained and substituted into the objective function formula to directly solve the minimum operating cost OF.

[0031] The verification process is as follows: Analyze and verify this embodiment with a 10 - unit (generator set) test system, show the adjustment process of the heuristic strategy based on the priority order method, and analyze the unit output, start - stop status, operating cost, and solution time respectively.

[0032] Calculate according to the formula and method provided in this embodiment. First, sort the 10 units from small to large using the sorting factor; secondly, obtain the initial solution on the basis of considering the system power balance constraint, spinning reserve constraint, and unit output boundary constraint; then, use the unit adjustment strategy based on the minimum start - stop time constraint, simple redundant unit adjustment strategy, and unit adjustment strategy based on the ramp - up / down constraint to adjust the unit output and start - stop status to obtain an approximate solution; finally, use the precise redundant unit adjustment strategy to adjust the unit output and start - stop status, perform the output adjustment between units to obtain the final solution, and calculate the minimum operating cost.

[0033] From Figure 2It can be seen that the order of the units after sorting is Unit 1, 2, 4, 3, 5, 6, 7, 8, 9, 10. The output curve of the initial solution satisfies the unit output boundary constraint, system power balance constraint, and spinning reserve constraint. However, some units violate the minimum start-stop time constraint and unit ramp rate constraint at some moments. From the start-stop status of the units, the minimum downtime of Unit 5 is 6 hours, but the downtime at the 17th moment is less than 5 hours. The minimum downtime of both Unit 6 and Unit 7 is 3 hours, but the downtime of Unit 6 at the 22nd moment and Unit 7 at the 22nd moment is less than 3 hours, violating the minimum start-stop time constraint. From the change of unit output, Units 1, 2, 8, 9, and 10 do not violate the unit ramp rate constraint. Unit 4 violates the unit ramp rate constraint at the 5th, 6th, 17th - 18th, and 23rd moments; Unit 3 violates the unit ramp rate constraint at the 6th - 7th, 16th, 18th - 19th, and 23rd moments; Unit 5 violates the unit ramp rate constraint at the 9th, 15th, 20th, and 22nd moments; Unit 6 violates the unit ramp rate constraint at the 10th, 14th, and 21st moments; Unit 7 violates the unit ramp rate constraint at the 12th - 13th moments. Through the analysis of the initial solution, it can be seen that the initial solution of the priority order method violates the constraints. Next, the proposed heuristic adjustment strategy will be used to adjust the unit output and start-stop status to meet the constraints of the unit commitment and improve the quality of the solution.

[0034] It can be seen from Figure 3 that the start-stop status of the approximate solution of the 10 units has changed. Each color square except red represents the corresponding unit being turned on. The red square indicates that the unit is restarted at this moment after adjustment, that is, the start-stop status of Units 5, 6, and 7 has changed. The minimum downtime of Unit 5 is 6 hours, but the downtime at the 17th moment is less than 5 hours, so it is turned on at the 16th and 17th moments. Similarly, Unit 6 is turned on at the 22nd moment and Unit 7 is turned on at the 22nd moment.

[0035] It can be seen from Figure 4It can be seen that after adjusting by using the unit adjustment strategy based on the minimum start-stop time constraint, the simple redundant unit adjustment strategy, and the unit adjustment strategy based on the ramping constraint, there is no obvious change in the output curves of units 1, 2, 8, 9, and 10 in the approximate solution. The output of unit 4 increases at the 3rd - 5th, 16th - 17th, and 22nd - 23rd moments, meeting the unit ramping constraint for the entire period. Similarly, the output of unit 3 increases at the 4th - 7th, 16th - 18th, and 22nd moments; the output of unit 5 increases at the 6th - 9th, 14th - 19th, and 21st - 22nd moments; the output of unit 6 increases at the 9th - 10th and 13th - 14th moments; the output of unit 7 increases at the 11th and 13th moments. Generally speaking, through the heuristic adjustment method of the strategy "increase in the forward direction and check in the reverse direction", the problem that the output of the unit fluctuates greatly and does not meet the unit ramping constraint is solved, making the change of the unit output smoother, and the ramping situation of the unit has been significantly improved.

[0036] It can be seen from Figure 5 that after adjusting by using the precise redundant unit adjustment strategy, except for the unchanged output of non - priority units 8, 9, and 10, the output of the remaining units has decreased. Among them, the output of unit 1 decreases at the 22nd moment; the output of unit 2 decreases at the 3rd - 5th, 16th - 19th, and 22nd - 23rd moments; the output of unit 4 decreases at the 5th - 9th, 17th - 19th, and 21st - 23rd moments; the output of unit 3 decreases at the 7th - 9th, 13th - 16th, 18th - 19th, and 21st - 22nd moments; the output of unit 5 decreases at the 9th - 11th and 13th - 16th moments; the output of unit 6 decreases at the 11th moment; the output of unit 7 decreases at the 9th - 11th and 13th - 16th moments. Generally speaking, the final solution meets the requirements of the unit ramping constraint and reduces the output of redundant units.

[0037] Table 1 Comparison of the operating results of 10 units

[0038] The test results obtained by the solution method proposed in this embodiment are compared with the test results obtained by the commercial solver Gurobi 11.0.1. It can be seen from Table 1 that the operating cost of the proposed solution method is about 0.59% higher than that of Gurobi, however, the solution time of this solution method is 0.46 s less than that of Gurobi, a reduction of nearly 40.7%. This shows that although the solution method proposed in this embodiment compromises in terms of accuracy, it does not require complex iterative calculations in terms of solution time, has a faster solution speed, and is suitable for solving the unit commitment problem with high time requirements.

Claims

1. A heuristic solution method for unit commitment based on priority order method, characterized by The following steps are involved: S1. With the goal of minimizing fuel cost and startup cost, the objective function of the power system unit combination model is constructed, and the unit output boundary constraints, system power balance constraints, spinning reserve constraints, unit ramp constraints, start-stop logic constraints and minimum start-stop time constraints are set; S2, using the average full load cost as the ranking factor, ranking the units in ascending order; S3, determining the start and stop states of the units to satisfy the spinning reserve constraint, and then determining the unit output to satisfy the system power balance constraint and the unit output boundary constraint, thereby obtaining an initial solution to the unit combination problem; S4, adjusting the unit output and start / stop status using a unit adjustment strategy based on minimum start / stop time constraints; S5. Set a simple redundant unit adjustment strategy to adjust the unit output and start / stop status, and obtain an approximate solution to the unit combination problem; S6. Use the unit adjustment strategy based on the ramp constraint to adjust the unit output and start / stop status; S7. Set accurate redundant unit adjustment strategy to adjust unit output and start / stop status; S8. Adjust the output between the units to obtain the final solution and calculate the minimum operating cost.

2. A heuristic solution method for unit commitment based on priority method according to claim 1, characterized in that: In S1, the objective function OF of the power system unit combination model is: ； Where N T is the total optimization period, t represents one of the moments; N G is the number of units, g represents one of the units; is the output of unit g at time t; , is a 0-1 integer variable, which represents the start and stop status of unit g at time t and time t-1 respectively. The value is 1 for start and 0 for stop; is the startup cost of unit g; is the fuel cost of unit g; The expression is: ； In the formula, , , is the power generation cost coefficient of unit g; The expression is: ； In the formula, , are the hot start cost and cold start cost of unit g respectively; is the continuous downtime time of unit g at time t; is the minimum downtime of unit g; is the cold start time of unit g.

3. The heuristic solution method for unit commitment based on priority method according to claim 2 is characterized in that: In S1, the constraints are specifically: Unit output boundary constraints: ； In the formula, , Respectively represent the minimum and maximum values ​​of the unit's g output; System power balance constraints: ； In the formula, represents the load at time t; Spinning reserve constraints: ； In the formula, represents the spare capacity at time t; Unit climbing constraints: ； ； In the formula, is the output of unit g at time t-1; , They represent the speed limits of the unit g power increase and decrease respectively; and They represent the startup power speed limit and shutdown power speed limit of unit g respectively; and All are 0-1 integer variables. Indicates the startup operation of unit g at time t. A value of 1 indicates that the startup operation was performed at time t, and a value of 0 indicates that the startup operation was not performed; Indicates the shutdown operation of unit g at time t. A value of 1 indicates that the shutdown operation was performed at time t, and a value of 0 indicates that the shutdown operation was not performed; Start and stop logic constraints: ； ； Minimum start and stop time constraints, including minimum startup time constraints and minimum shutdown time constraints: ； ； In the formula, , represents a 0-1 integer variable, which represents the start-up and shutdown operations of unit g at time s; is the minimum startup time of unit g; Indicates a universal quantifier.

4. The heuristic solution method for unit commitment based on priority method according to claim 3 is characterized in that: In S2, the average full-load cost refers to the average power generation cost of the unit in full power state, which is numerically the power generation cost of the unit in full power state divided by the maximum output of the unit. For unit g, its average full-load cost is It is expressed as: 。 5. The heuristic solution method for unit commitment based on priority method according to claim 3 is characterized in that: In S3, the process of obtaining the initial solution of the unit combination is: S3.

1. Determine the start and stop status of the units to meet the spinning reserve constraint. Assuming that the start and stop status of all units at each moment is 0, the process is as follows: S3.1.1, let t = 1; S3.1.2, let g = 1, indicating the first unit after sorting; S3.1.

3. Order , It is the sum of the maximum output of the units started at time t; S3.1.

4. Order ; S3.1.5 If ,but , let g=g+1, if at this time , go to step S3.1.4, otherwise go to step S3.1.6, where Indicates the last unit after sorting; like ,but , go to step S3.1.6; S3.1.6, let t = t + 1, if , go to step S3.1.2, otherwise end the adjustment; S3.

2. Determine the unit output to satisfy the system power balance constraint and the unit output boundary constraint, thereby obtaining the initial solution to the unit combination problem. Assuming that the output of all units at each moment is 0, the process is as follows: S3.2.1, let t = 1; S3.2.2, let g = 1; S3.2.

3. Order , is the sum of the power of the units started at time t; S3.2.4 If , go to step S3.2.6, otherwise , at this time, if ,but ,like ,but ; S3.2.5, let g = g + 1, if , go to step S3.2.4, otherwise go to step S3.2.6; S3.2.6, let t = t + 1, if , go to step S3.2.2, otherwise end the adjustment.

6. A heuristic solution method for unit commitment based on priority method according to claim 5, characterized in that: In the above S4, the unit output and start-stop state are adjusted by using the unit adjustment strategy based on the minimum start-stop time constraint. Specifically, if the minimum start-up time constraint cannot be met at a certain moment, the unit is started at the next moment; if the minimum stop time constraint cannot be met at a certain moment, the unit is started at that moment; when the state of unit g at time t changes from 0 to 1, the output of unit g is .

7. The heuristic solution method for unit commitment based on priority order method according to claim 6 is characterized in that: In the above S5, a simple redundant unit adjustment strategy is set to adjust the unit output and start-stop state, and an approximate solution to the unit combination problem is obtained. Specifically, when the sum of the unit output exceeds the total load demand, the system output is redundant, and the redundant unit output needs to be adjusted at this time; during the adjustment process, the simple redundant unit adjustment strategy takes into account the influence of the unit output boundary constraint, the system power balance constraint and the spinning reserve constraint, and does not involve the unit climbing constraint; the adjustment starts from the unit with the lowest priority, and the process is as follows: S5.1.1, let t = 1; S5.1.

2. Calculate the redundant output at time t : ； S5.1.

3. Order ; S5.1.

4. Order , , Indicates the adjustable amount of the unit g, is the total adjustment amount at time t; S5.1.5 Calculation : ； S5.1.6 If ,but , ; Let g = g-1. If g ≥ 1, go to step S5.1.

5. Otherwise, go to step S5.1.

7. like ,but , go to step S5.1.7; S5.1.7, let t = t + 1, if , go to step S5.1.2, otherwise end the adjustment.

8. The heuristic solution method for unit commitment based on priority order method according to claim 7 is characterized in that: In the above S6, the unit output and start-stop status are adjusted by using the unit adjustment strategy based on the climbing constraint, specifically, the time periods and units are classified and processed according to the start-stop changes of the units; the unit continuous start-up period is recorded as a "1" type period, and the unit continuous shutdown period is recorded as a "0" type period; The units that remain on all the time are called normally-on units, the units that remain off all the time are called normally-off units, and the units that change between starting and stopping are called start-stop units. Among them, normally-off units and normally-on units that maintain maximum output all the time always meet the unit climbing constraints, so no adjustment is required. For the normally-on units, the output of the normally-on units with output changes during the whole period is adjusted to meet the unit climbing constraints. Since the whole period of this type of unit is a "1" type period, the method of "increasing in the forward direction and testing in the reverse direction" is adopted to adjust the unit output to meet the unit climbing constraints. The specific process is as follows: S6.1.

1. Increase in time, the steps include: S6.1.1.

1. Order ,in is the initial moment of the "1" type period; S6.1.1.2 If ,but ; like ,but ; S6.1.1.3, let t = t + 1, if , go to step S6.1.1.2, otherwise end the adjustment, where is the end time of the "1" type period; S6.1.2, reverse time test, the steps include: S6.1.2.

1. Order ; S6.1.2.2 If ,but ; S6.1.2.3, let t = t-1, if , go to step S6.1.2.2, otherwise end the adjustment; For the start-stop units, the output and start-stop status of the start-stop units are adjusted to meet the unit climbing constraints; the full period of the start-stop units includes both the "1" type period and the "0" type period; first, the "1" type period is adjusted by the method of "increase in time and check in reverse time" to meet the unit climbing constraints; then the start-stop units are adjusted for the "0" type period containing different times. The specific process is as follows: S6.2.

1. For the first end time, when the "0" type period contains the first end time, it is necessary to adjust the unit output and start / stop status at the right end time of the "0" type period; assuming that the "0" type period is "1 to t1", t1 is the right end time of the "0" type period, then: S6.2.1.1, let t = t1; S6.2.1.2 If ,but and , and go to step S6.2.1.3, otherwise end the adjustment, where, is the output of unit g at time t+1; S6.2.1.3, let t = t-1, if t ≥ 1, go to step S6.2.1.2, otherwise end the adjustment; S6.2.2 For the terminal moment, when the "0" type period contains the terminal moment, it is necessary to adjust the unit output and start / stop status at the left end of the "0" type period; assuming that the "0" type period is "t2 to N T ", t2 is the left end time of the "0" type period, then: S6.2.2.1, let t = t2; S6.2.2.2 If ,but and , and go to step S6.2.2.3, otherwise end the adjustment, where, is the output of unit g at time t-1; S6.2.2.3, let t = t + 1, if t ≤ N T , go to step S6.2.2.2, otherwise end the adjustment; S6.2.

3. For the intermediate moments, when the "0" type period only includes the intermediate moments, it is necessary to adjust the unit output and start / stop status at both ends of the "0" type period; assuming that the "0" type period is "t3 to t4", t3 and t4 are the left and right end moments of the "0" type period, respectively, then: S6.2.3.

1. For the left adjustment of the "0" type period, let t = t3; S6.2.3.2 If ,but and , go to step S6.2.3.3, otherwise end the left end adjustment; S6.2.3.3, let t = t + 1, if t ≤ t4, go to step S6.2.3.2, otherwise end the left end adjustment; S6.2.3.4, for the right end adjustment of the "0" type period, let t = t4; S6.2.3.5: If ,but and , go to step S6.2.3.6, otherwise end the right end adjustment; S6.2.3.6, let t = t-1, if t ≥ t3, go to step S6.2.3.5, otherwise end the right end adjustment; S6.2.3.

7. After the adjustment at both ends is completed, the downtime of the unit during the "0" period is counted. ;like , indicating that the "0" type period still meets the minimum downtime constraint, and the adjustment is completed; if , indicating that the "0" type period cannot meet the minimum downtime constraint; in order to maintain the system power balance, the "0" type period and the previous and next "1" type periods are combined into a "1" type period; finally, the "forward-time increase and reverse-time inspection" method is adopted to adjust the unit output of the newly formed "1" type period to meet the unit climbing constraint.

9. The heuristic solution method for unit commitment based on priority order method according to claim 8, characterized in that: In the above S7, the precise redundant unit adjustment strategy is used to adjust the unit output and start-stop status at each moment in turn on the basis of satisfying various constraints, so as to obtain the final solution to the unit combination problem. The specific process is as follows: since the unit output at the head end moment and the tail end moment is affected by the unit climbing constraint at one end adjacent moment, and the middle moment is affected by the unit climbing constraint at both ends adjacent moment, the adjustment process divides the whole period into the head end period including the head end moment, the middle period including only the middle moment, and the tail period including the tail moment for redundant unit adjustment; For the headend period: S7.1.1, let t = 1; S7.1.

2. Calculate the redundant output R1 at the head end: ； In the formula, Indicates the output of unit g at the first end moment; Indicates the load at the head end; S7.1.

3. Order ; S7.1.

4. Order and ; S7.1.5, Computer group g adjustable amount: ； In the formula, Indicates the output of unit g at the right end of the first end period; S7.1.6 If ,but and , let g = g-1, if g ≥ 1, go to step S7.1.5, otherwise end the adjustment; like ,but , end the adjustment; For the middle period: S7.2.1, let t = 2; S7.2.

2. Calculate the redundant output at time t: ； S7.2.

3. Order ; S7.2.

4. Order and ; S7.2.5, Computer group g adjustable amount: ； S7.2.6 If ,but and , let g = g-1, if g ≥ 1, go to step S7.2.5, otherwise go to step S7.2.7; like ,but , go to step S7.2.7; S7.2.7, let t = t + 1, if , then go to step S7.2.2, otherwise end the adjustment; For the terminal period, the adjustment process of the terminal period is the same as that of the head period. Only the influence of the output of the unit at one end needs to be considered, that is, the output of the unit at the left end of the terminal period.

10. The heuristic solution method for unit commitment based on priority order method according to claim 9, characterized in that: In the above S8, after the adjustment of the redundant units in the head-end period, the middle period and the end period is completed, the sum of the unit outputs at each moment is fixed. At this time, it is necessary to adjust the outputs between the units. The adjustment method is to divide the entire period into the head-end period, the middle period and the end period, and adjust by increasing the output of the priority unit and reducing the output of the non-priority unit; wherein, the adjustment process of the middle period is as follows: S8.1, let t = 2; S8.2, let g = 1; S8.3 If , go to step S8.9, otherwise: ； In the formula, Indicates the power that the priority unit can increase; S8.

4. Order , , Indicates the power that non-priority units can reduce; Indicates the total reduction in power of non-priority units; S8.

5. Order , i is the index of the unit; S8.6 If ,but: ； Otherwise, go to step S8.8; in, is a 0-1 integer variable, indicating the start / stop status of unit i at time t; , , Respectively represent the output of unit i at time t, time t-1, and time t+1; , They represent the speed limits for the increase and decrease of the power of unit i respectively; is the minimum output of unit i; S8.7, if ,but: ； ； ； Go to step S8.8; like ,but: ； ； Go to step S8.10; S8.8, let i = i-1, if i>g, go to step S8.6, otherwise go to step S8.9; S8.9, let g = g + 1, if , go to step S8.3, otherwise go to step S8.10; S8.10, let t = t + 1, if , go to step S8.2, otherwise end the adjustment; For the first and last time periods, they are only affected by the ramp constraints of the units at adjacent times at one end, and can be adjusted at one end in the same way as the adjustment process of the middle time period; After the adjustment is completed, the minimum operating cost of the computer group combination problem is calculated based on the obtained unit output and start-stop status.

Images