Runoff hydropower station annual maintenance and power generation plan joint optimization method based on improved ant colony algorithm

Abstract

The invention provides a runoff hydropower station annual maintenance and power generation plan joint optimization method based on an improved ant colony algorithm, and belongs to the technical field of hydropower station optimization scheduling. According to the method, for the multi-target conflict problem of abandoned water loss, peak regulation capacity insufficiency and base load vacancy in an annual maintenance plan of a runoff hydropower station, a comprehensive target function containing abandoned water penalty, peak regulation vacancy penalty and base load vacancy penalty is constructed, and combined optimization is achieved by adopting an improved multi-group coevolution ant colony algorithm. The core of the algorithm lies in that ant colonies are divided into common ant colonies and elite ant colonies, the search capability is improved through local and global pheromone updating and diffusion mechanisms, and the process is optimized to determine the optimal overhaul time period of a to-be-overhauled unit and the optimal operation mode of a non-overhaul unit. Example verification shows that the method can significantly reduce abandoned water loss and base load vacancy, ensures sufficient peak regulation capacity, saves annual operation cost of the hydropower station, and is suitable for intelligent decision-making of the large radial flow type hydropower station.

Description

Technical Field

[0001] This invention belongs to the field of hydropower station optimization scheduling technology, specifically a method for joint optimization of annual maintenance and power generation plans of run-of-river hydropower stations based on an improved ant colony algorithm. Background Technology

[0002] Existing hydropower station optimization scheduling technologies mainly include: [1]ABIRAMI M, GANESAN S,SUBRAMANIAN S, et al. Source and transmission line maintenance outage scheduling in a power system using teaching learning based optimization algorithm [. Applied SoftComputing, 2014, 21:72-83. [2] Zhou Ming, Xia Shu, Li Yan, et al. Joint optimization scheduling of monthly unit combination and maintenance plan for power systems including wind power [J]. Proceedings of the CSEE, 2015, 35(7): 1586-1595. Zhou Ming, Xia Shu, Li Yan, et al. A joint optimization approachonmonthly unit commitment and maintenance scheduling for windpower integrated power systems[J]. Proceedings of the CSEE, 2015,35(7):1586-1595. [3]Froger A, Gendreau M, Mendoza JE, et al. Maintenance scheduling in the electricity industry: a literature review[I]. European Journal of Operational Research, 2016,251(3):695-706. [4] Su Chengguo, Wang Peilin, Wu Xinyu, et al. MILP model for short-term peak shaving of cascade hydropower stations considering unit combination [J]. Power System Technology, 2018, 42(6):1883-1891. Su Chengguo, Wang Peilin, Wu Xinyu, et al. A compact MILP model for short-term peak shaving of cascaded hydropower plants considering unitcommitment[J]. Power System Technology, 2018,42(6): 1883-1891. [5] Meng Tao, Pan Jie, Hu Jingwen, Research on power outage maintenance plan under the power market [J]. China Electric Power Enterprise Management, 2020(27):44-46. MENG Tao, PAN Jie, HU Jingwen, Research on outage maintenance plan inpower market[]]. China Power Enterprise Management, 2020(27):44-46. [6] Ji Changming, Li Chuangang, Liu Xiaoyong, et al. Research on dynamic programming algorithm based on functional analysis and its application in reservoir scheduling [J]. Journal of Hydraulic Engineering, 2016, 47(1):1-9. Ji Changming, Li Chuangang, Liu Xiaoyong, etal. Research and application of dynamic programming algorithm in reservoir operation based on functional analysis[1]. Journal of Hydraulic Engineering, 2016,47(1): 1-9. [7]El-Sharkh M Y. Clonal selection algorithm for powergeneratorsmaintenance scheduling[. International Journal of Electrical Power&EnergySystems,2014,57(3):73-78. The existing technology has significant shortcomings: (1) Multi-objective conflict problem: The annual maintenance of run-of-river hydropower stations faces the contradiction of water abandonment loss (flood season water fluctuation), insufficient peak-shaving capacity (grid demand) and base load deficit (accounting for 40%-60% of the annual power generation), which is difficult to coordinate with traditional manual scheduling.

[0003] (2) Algorithm defects: Domestic and foreign research generally focuses on single-objective or dual-objective optimization, such as mixed integer programming (references [1-2], which ignores the coupling effect of water abandonment and peak shaving; genetic algorithm and particle swarm optimization (references [3-4]) are prone to premature convergence in high-dimensional solution space; stochastic programming (references [5-6]) fails to accurately handle the characteristics of steep rise and fall in runoff. More importantly, existing methods ignore the base load guarantee requirements, resulting in excessive base load deficit during maintenance.

[0004] (3) Practical application limitations: Conventional ant colony algorithms suffer from premature convergence, the curse of dimensionality (the solution space reaches 365N when the number of units to be repaired is >10), and the difficulty of coordinating the weights of multiple targets with a single pheromone matrix.

[0005] Therefore, a new optimization method is urgently needed to achieve multi-objective synergistic advantages. Summary of the Invention

[0006] In view of this, the present invention provides a joint optimization method for annual maintenance and power generation plans of run-of-river hydropower stations based on an improved ant colony algorithm. This method is used for intelligent decision-making in large-scale run-of-river hydropower stations to solve the multi-objective conflict problem of water abandonment loss, insufficient peak-shaving capacity and base load deficit. The improved ant colony algorithm achieves efficient optimization, thereby improving the economic benefits of hydropower stations and the stability of the power grid.

[0007] To achieve the above-mentioned technical features, the objective of this invention is as follows: a joint optimization method for annual maintenance and power generation plans of run-of-river hydropower stations based on an improved ant colony algorithm, the method comprising: Step 1: Construct a multi-objective optimization model; A comprehensive objective function is constructed with the goal of minimizing annual water wastage losses, peak load deficits, and base load deficits. Step 2: Based on the actual power generation curves of the hydropower station over the years, the frequency analysis method is used to decompose the base load component and peak load. Step 3, define virtual nodes and virtual edges, virtual nodes U m,t Indicates the unit m At the initial time t Initial maintenance decision, virtual edge e m1 , e m2 Indicates from node U m1,t1 arrive U m2,t2 The transfer operation; Step 4: Optimize the maintenance period and operation mode using an improved multi-population co-evolutionary ant colony algorithm; the optimization process includes initialization, ant path selection, and local and global pheromone updates. Step 5: Output the optimal maintenance plan and power generation scheme to ensure that the annual water wastage loss, peak load deficit and base load deficit are minimized.

[0008] Preferably, the objective function constructed in step 1 is: (1) In the formula: d For date, The amount of electricity converted from the daily water wastage. This is the penalty coefficient for water abandonment. This refers to the daily peak-shaving shortfall in electricity volume. This is the peak-shaving penalty coefficient. This refers to the daily baseload deficit. This is the baseload shortfall penalty coefficient.

[0009] Preferably, in the objective function construction process in step 1, the three objectives are quantitatively balanced by normalized weighted water wastage penalty, peak load deficit penalty, and baseload deficit penalty.

[0010] Preferably, the improved multi-population co-evolutionary ant colony algorithm in step 4 includes the following steps: Step 4.1: Divide the ant colony into ordinary ant colonies and elite ant colonies; Step 4.2: The ordinary ant colony generates a new solution based on the Gaussian function: Description of the Gaussian function for ordinary ants: (2) (3) (4) In the formula, It is the first Gaussian function of a common ant To optimize the solution to the problem, It is the first The generation The mean of a normal distribution, It is the first The generation The standard deviation of a normal distribution It is the first The generation The variance of a normal distribution, It is the first In the distribution, the th... The fitness value of each sample It is the first Scaling factor of the generation, Population size; Step 4.3: The elite ant colony uses a probabilistic selection method combined with a Gaussian kernel function to generate a solution; Step 4.4 introduces a pheromone diffusion mechanism to improve global search capabilities and avoid premature convergence; Step 4.5: Adaptively adjust the global pheromone attenuation coefficient.

[0011] Preferably, in step 4.1, elite ant colonies account for 10% and ordinary ant colonies account for 90%. Preferably, the pheromone diffusion mechanism introduced in step 4.4 specifically includes: The local pheromone update formula is as shown in equation (5), which updates the maintenance node. To the maintenance node Pheromones between them: (5) In the formula, Let be the number of iterations of the ant colony. Indicates the rate at which an individual's pheromones evaporate. The range of values ​​is ∈ (0,1), Indicates the local pheromone residue coefficient. For ants from the maintenance node To the maintenance node The increase in pheromones between them For the current iteration N At this time, node to node The concentration of pheromones along the path between them After the update The next iteration; The global pheromone update formulas are as shown in equations (6) and (7): (6) (7) According to equations (6) and (7), This indicates after the update, i.e. Global pheromone levels after the next iteration; It is the present moment, that is Pheromones concentration after the next iteration It is the global pheromone evaporation rate; It is the global pheromone growth coefficient. The current objective function value, if It was before The minimum value in the next iteration, i.e., the optimal value, is the pheromone increment. for If not, then It is 0.

[0012] Preferably, in step 4.5, the global pheromone attenuation coefficient is adaptively adjusted as shown in equation (8): (8) In the formula, for Pheromones attenuation coefficient, and These are the upper and lower limits of pheromone attenuation. The optimal value for the current iteration objective. This represents the optimal value for the objective across all historical iterations.

[0013] Preferably, it further includes a constraint processing step: 1) Continuity constraints for unit maintenance, as shown in equation (9): (9) In the formula, For unit maintenance status variables, when the unit In the d During daily maintenance, its value is 1; otherwise, its value is 0. For the unit Maintenance start date. For the unit Maintenance duration; For hydropower unit maintenance, the unit maintenance must be completed continuously in one go to avoid interruption leading to reduced maintenance efficiency or equipment damage risk. 2) Constraints for synchronous maintenance of generating units, as shown in formula (10): (10) In the formula, A collection of all units awaiting maintenance. This is the maximum number of generating units that can be maintained during the same period; run-of-river hydropower stations need to maintain enough operating units to meet basic power generation and peak load requirements; if multiple units are under maintenance at the same time, it may lead to insufficient power output of the power station, exacerbating water wastage or peak load / base load power shortages. 3) Unit start-up and shutdown constraints, as shown in equations (11)-(13): (11) (12) (13) In equations (11)-(13), For the unit m In the d Daytime t The start / stop status variable has a value of 1 when it is in the power-on state and a value of 0 otherwise. , These represent the duration of continuous unit startup / shutdown. 4) Water balance constraint, as shown in equation (14): (14) As shown in equation (14). and The respective d sky t Time period and t- Storage capacity for one time period; This represents the time step; the constant 3600 represents 1. h Equals 3600s. and The first d Daytime t The inflow and outflow of the hydropower station m 3 / s ; 5) Outbound flow constraints, as shown in equations (15) and (16): (15) (16) As shown in equations (15)-(16), t Outbound flow during different time periods Power generation flow of the power plant and water discharge flow sum; and These are the maximum and minimum outbound flow limits, respectively. 6) Output constraints of hydropower units, as shown in equations (17) and (18): (17) (18) In the formula, NHQ represents the dynamic characteristics of the unit, and represents the nonlinear functional relationship between the unit's output and the water head and power generation flow rate. The generating flow rate of the unit is m 3 / s, For the generator head of the unit, m ; , For the unit m Maximum / minimum technical output limits; 7) Gradient constraint of hydropower unit, as shown in equation (19): (19) In the formula, and Hydropower station unitsm Time period t Internal output limits for downward and upward climbing; 8) Vibration zone constraint, as shown in equation (20): (20) In the formula, , For the unit m The k The upper and lower limits of output in each vibration zone; 9) Generating head constraint of the unit, as shown in equation (21): ;(twenty one) In the formula, For the unit m In the d Daytime t Hydropower head; For the first d Daytime t The reservoir water level; For the first d Daytime t The tailwater level; 10) Reservoir water level constraints, as shown in equations (22) and (23): ;(twenty two) ;(twenty three) In the formula, For the reservoir water level, and The first d Minimum / maximum water level limits for the day This is the maximum allowable fluctuation in water level for that day. 11) Water level-reservoir capacity constraint, as shown in equation (24): ;(twenty four) In the formula, This is the water level-reservoir capacity curve function of the power station. For storage capacity; 12) Tailwater level - outflow constraint, as shown in equation (25): (25) In the formula, This is a function of the tailrace level and outflow rate of the power station. 13) Upper and lower peak-shaving capacity constraints: (26) (27) In the formula, For the unitm In the d Heavenly t Peak capacity adjustment during specific time periods Ramp-up restrictions for unit peak shaving For the unit m In the d Heavenly t The operational status of a time period is represented by a binary variable, where 1 indicates operation and 0 indicates shutdown. For the unit m The upper limit of output, For the unit m In the d Heavenly t Actual power generation during the period Non-negative constraints for peak capacity adjustment; For the unit m In the d Heavenly t Peak-shaving capacity during the period, Ramp-up restrictions for unit peak shaving For the unit m The lower limit of output, Non-negative constraint for peak reduction capacity; As shown in equations (26) and (27), the operating generating units reserve up and down adjustment capabilities to meet the peak shaving needs of the power grid.

[0014] Another aspect of the present invention provides a joint optimization system for annual maintenance and power generation plans of run-of-river hydropower stations based on an improved ant colony algorithm. The system is used in the aforementioned joint optimization method for annual maintenance and power generation plans of run-of-river hydropower stations based on an improved ant colony algorithm, and includes: Data input module: used to import historical power generation curves of hydropower stations, unit parameters, inflow data and grid peak shaving requirements; Optimization Engine Module: Performs joint optimization based on the improved ant colony algorithm, including collaborative search, pheromone update and diffusion between ordinary ant colonies and elite ant colonies; Simulation calculation module: Implements objective function calculation and constraint verification in the MATLAB environment; Output module: Generates annual maintenance plan and power generation operation mode diagram; Preferably, the unit parameters include installed capacity and maintenance cycle.

[0015] The present invention has the following beneficial effects: 1. This invention constructs a comprehensive objective function through a multi-objective optimization framework for peak shaving and baseload shortage, and normalizes the weighted water abandonment penalty, peak shaving shortage penalty, and baseload shortage penalty to achieve a quantitative trade-off among the three objectives.

[0016] 2. This invention improves the ant colony algorithm by adopting a multi-population co-evolution mechanism, dividing the ant colony into ordinary ant colonies and elite ant colonies. Ordinary ant colonies use a Gaussian function to enhance the diversity of local search; elite ant colonies optimize the convergence speed based on probabilistic selection and introduce a pheromone diffusion mechanism to improve global search capabilities and avoid premature convergence.

[0017] 3. This invention employs a dual-population collaborative architecture, with one set optimizing maintenance timing (i.e., virtual node definition) and the other set optimizing unit combination strategy (i.e., operating mode), to process the high-dimensional solution space through co-evolution.

[0018] 4. This invention is supported by a mathematical model: defining a complete set of constraints, including maintenance continuity, water balance and unit output constraints, to ensure the feasibility of the solution.

[0019] 5. Through case studies, this invention demonstrates that it can reduce water wastage losses, ensure peak-shaving capacity, eliminate base load deficits, and reduce total operating costs by 15.4%-24.1%, which is significantly better than traditional algorithms.

[0020] In summary, this invention is applicable to intelligent decision-making for large run-of-river hydropower stations (such as power stations without regulating reservoir capacity), solving the multi-objective conflict problem of water abandonment loss, insufficient peak-shaving capacity and base load deficit. It achieves efficient optimization through improved ant colony algorithm, thereby improving the economic benefits of hydropower stations and the stability of the power grid. Attached Figure Description

[0021] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0022] Figure 1 Schematic diagram of the daily regulation capacity of a run-of-river hydropower station.

[0023] The daily load curve is shown to be decomposed into the base load curve and peak load capacity, explaining the principle that improper maintenance planning leads to peak load deficit (tp3-tp4 interval) and base load deficit (tp2 interval).

[0024] Figure 2 Flowchart of the improved ant colony algorithm.

[0025] The algorithm's solution steps are as follows: initialize ant paths, select local pheromone updates, global pheromone updates, ant colony evolution (selection, crossover, mutation), adaptive adjustment, and output the optimal solution.

[0026] Figure 3 Annual overhaul plan for generating units based on improved ant colony algorithm.

[0027] Display the optimal maintenance schedule (dry season, January to June) to avoid maintenance during the flood season (July to August) and ensure efficient water utilization.

[0028] Figure 4 Output and base load curves of units undergoing maintenance.

[0029] By comparing actual output with baseload demand, it is proven that the optimal solution ensures zero deviation in baseload supply.

[0030] Figure 5 The peak-shaving potential of hydropower stations considering maintenance is shown by the peak-shaving capacity curve of the operating units, indicating that the plan meets the peak-shaving needs of the power grid (except for local periods).

[0031] Figure 6 Considering the discharge of water from hydropower stations undergoing maintenance.

[0032] Describe the distribution of water release throughout the year, highlighting the concentrated water release during the flood season.

[0033] Figure 7 Annual maintenance plan for generating units based on conventional ant colony algorithm.

[0034] As a comparison, it shows that the maintenance schedule is unreasonable (such as maintenance in January and February leading to a shortage of peak load).

[0035] Figure 8 Annual overhaul plan for generating units based on particle swarm optimization algorithm.

[0036] Another alternative approach involves focusing maintenance during the dry season, but neglecting peak load periods, thus exacerbating the base load deficit. Detailed Implementation

[0037] The embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0038] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0039] Example 1: Step 1: Construct a multi-objective optimization model; To minimize annual water wastage losses, peak load deficits, and base load deficits, a comprehensive objective function is constructed: (1) In the formula: d For date, The amount of electricity converted from the daily water wastage. This is the penalty coefficient for water abandonment. This refers to the daily peak-shaving shortfall in electricity volume. This is the peak-shaving penalty coefficient. This refers to the daily baseload deficit. This is the baseload shortfall penalty coefficient.

[0040] Step 2: Based on the actual power generation curves of the hydropower station over the years, the frequency analysis method is used to decompose the base load component and peak load. Step 3, define virtual nodes and virtual edges, virtual nodes U m,t Indicates the unit m At the initial time t Initial maintenance decision, virtual edge e m1 , e m2 Indicates from node U m1,t1 arrive U m2,t2 The transfer operation; Step 4: Optimize the maintenance period and operation mode using an improved multi-population co-evolutionary ant colony algorithm; the optimization process includes initialization, ant path selection, and local and global pheromone updates. Step 4.1: Divide the ant colony into ordinary ant colonies and elite ant colonies, with elite ant colonies accounting for 10% and ordinary ant colonies accounting for 90%. Step 4.2: The ordinary ant colony generates a new solution based on the Gaussian function: To prevent itself from reaching local optima and generating new solutions more slowly, common ants use a Gaussian function and the average value of each dimension. Equations (2) to (4) describe the Gaussian function of common ants.

[0041] Description of the Gaussian function for ordinary ants: (2) (3) (4) In the formula, It is the first Gaussian function of a common ant To optimize the solution to the problem, It is the first The generation The mean of a normal distribution, It is the first The generation The standard deviation of a normal distribution It is the first The generation The variance of a normal distribution, It is the first In the distribution, the th... The fitness value of each sample It is the first Scaling factor of the generation, Population size; Step 4.3: The elite ant colony uses a probabilistic selection method combined with a Gaussian kernel function to generate a solution; Step 4.4 introduces a pheromone diffusion mechanism to improve global search capabilities and avoid premature convergence; The local pheromone update formula is as shown in equation (5), which updates the maintenance node. To the maintenance node Pheromones between them: (5) In the formula, Let be the number of iterations of the ant colony. Indicates the rate at which an individual's pheromones evaporate. The range of values ​​is ∈ (0,1), Indicates the local pheromone residue coefficient. For ants from the maintenance node To the maintenance node The increase in pheromones between them For the current iteration N At this time, node to node The concentration of pheromones along the path between them After the update The next iteration; The global pheromone update formulas are as shown in equations (6) and (7): (6) (7) According to equations (6) and (7), This indicates after the update, i.e. Global pheromone levels after the next iteration; It is the present moment, that is Pheromones concentration after the next iteration It is the global pheromone evaporation rate; It is the global pheromone growth coefficient. The current objective function value, if It was before The minimum value in the next iteration, i.e., the optimal value, is the pheromone increment. for If not, then It is 0.

[0042] Step 4.5, adaptively adjust the global pheromone attenuation coefficient as shown in equation (8): (8) In the formula, for Pheromones attenuation coefficient, and These are the upper and lower limits of pheromone attenuation. The optimal value for the current iteration objective. This represents the optimal value for the objective across all historical iterations.

[0043] Step 5: Output the optimal maintenance plan and power generation scheme to ensure that the annual water wastage loss, peak load deficit and base load deficit are minimized.

[0044] Furthermore, it also includes constraint handling steps: 1) Continuity constraints for unit maintenance, as shown in equation (9): (9) In the formula, For unit maintenance status variables, when the unit In the d During daily maintenance, its value is 1; otherwise, its value is 0. For the unit Maintenance start date. For the unit Maintenance duration; For hydropower unit maintenance, the unit maintenance must be completed continuously in one go to avoid interruption leading to reduced maintenance efficiency or equipment damage risk. 2) Constraints for synchronous maintenance of generating units, as shown in formula (10): (10) In the formula, A collection of all units awaiting maintenance. This is the maximum number of generating units that can be maintained during the same period; run-of-river hydropower stations need to maintain enough operating units to meet basic power generation and peak load requirements; if multiple units are under maintenance at the same time, it may lead to insufficient power output of the power station, exacerbating water wastage or peak load / base load power shortages. 3) Unit start-up and shutdown constraints, as shown in equations (11)-(13): (11) (12) (13) In equations (11)-(13), For the unit m In the d Daytime t The start / stop status variable has a value of 1 when it is in the power-on state and a value of 0 otherwise. , These represent the duration of continuous unit startup / shutdown. 4) Water balance constraint, as shown in equation (14): (14) As shown in equation (14). and The respective d sky t Time period and t- Storage capacity for one time period; This represents the time step; the constant 3600 represents 1. h Equals 3600s. and The first d Daytime t The inflow and outflow of the hydropower station m 3 / s ; 5) Outbound flow constraints, as shown in equations (15) and (16): (15) (16) As shown in equations (15)-(16), t Outbound flow during different time periods Power generation flow of the power plant and water discharge flow sum; and These are the maximum and minimum outbound flow limits, respectively. 6) Output constraints of hydropower units, as shown in equations (17) and (18): (17) (18) In the formula, NHQ represents the dynamic characteristics of the unit, and represents the nonlinear functional relationship between the unit's output and the water head and power generation flow rate. The generating flow rate of the unit is m 3 / s, For the generator head of the unit, m ; , For the unit m Maximum / minimum technical output limits; 7) Gradient constraint of hydropower unit, as shown in equation (19): (19) In the formula, and Hydropower station units m Time period t Internal output limits for downward and upward climbing; 8) Vibration zone constraint, as shown in equation (20): (20) In the formula, , For the unit m The k The upper and lower limits of output in each vibration zone; 9) Generating head constraint of the unit, as shown in equation (21): ;(twenty one) In the formula, For the unit m In the d Daytime t Hydropower head; For the first d Daytime t The reservoir water level; For the first d Daytime t The tailwater level; 10) Reservoir water level constraints, as shown in equations (22) and (23): ;(twenty two) ;(twenty three) In the formula, For the reservoir water level, and The first d Minimum / maximum water level limits for the day This is the maximum allowable fluctuation in water level for that day. 11) Water level-reservoir capacity constraint, as shown in equation (24): ;(twenty four) In the formula, This is the water level-reservoir capacity curve function of the power station. For storage capacity; 12) Tailwater level - outflow constraint, as shown in equation (25): (25) In the formula, This is a function of the tailrace level and outflow rate of the power station. 13) Upper and lower peak-shaving capacity constraints: (26) (27) In the formula, For the unit m In the d Heavenly t Peak capacity adjustment during specific time periods Ramp-up restrictions for unit peak shaving For the unit m In thed Heavenly t The operational status of a time period is represented by a binary variable, where 1 indicates operation and 0 indicates shutdown. For the unit m The upper limit of output, For the unit m In the d Heavenly t Actual power generation during the period Non-negative constraints for peak capacity adjustment; For the unit m In the d Heavenly t Peak-shaving capacity during the period, Ramp-up restrictions for unit peak shaving For the unit m The lower limit of output, Non-negative constraint for peak reduction capacity; As shown in equations (26) and (27), the operating generating units reserve up and down adjustment capabilities to meet the peak shaving needs of the power grid.

[0045] Example 2: In another aspect, this invention provides a joint optimization system for annual maintenance and power generation plans of run-of-river hydropower stations based on an improved ant colony algorithm. The system is used to implement the aforementioned joint optimization method for annual maintenance and power generation plans of run-of-river hydropower stations based on an improved ant colony algorithm, comprising: Data input module: used to import historical power generation curves of hydropower stations, unit parameters, inflow data and grid peak shaving requirements; Optimization Engine Module: Performs joint optimization based on the improved ant colony algorithm, including collaborative search, pheromone update and diffusion between ordinary ant colonies and elite ant colonies; Simulation calculation module: Implements objective function calculation and constraint verification in the MATLAB environment; Output module: Generates annual maintenance plan and power generation operation mode diagram; Preferably, the unit parameters include installed capacity and maintenance cycle.

[0046] Example 3: The verification and application of the optimization results include: (1) Case verification: For a hydropower station with a total installed capacity of 3190W and 21 generating units (2 Class A 170W units and 19 Class B 150W units), 12 units are to be overhauled, the overhaul period is 14 days, and the maximum number of units to be overhauled at the same time is 2.

[0047] (2) Comparative analysis: Compared with conventional ant colony algorithm and particle swarm algorithm, the objective function value (total penalty cost) is reduced by 15.4%-24.1%.

[0048] (3) Effects: The maintenance plan is scheduled during the dry season (January-June) to avoid water wastage during the flood season; it ensures zero deviation in base load supply and has sufficient peak-shaving potential.

[0049] (4) Parameter settings: ant colony size m = 50, pheromone factor α = 1, heuristic function factor β = 2, maximum number of iterations gmax -200.

[0050] Example 4: Figure 1 A schematic diagram of the daily regulation capacity of a run-of-river hydropower station is provided, along with the daily load curve. The power generation plan curve of the hydropower station can be decomposed into a base load curve and a peak-shaving capacity. If the hydropower station's maintenance plan is reasonably arranged, the available capacity should meet the unit output regulation during peak load periods and reserve some standby capacity. If the capacity of the units to be maintained... P m ≤ P s - P max1 , P s For the installed capacity of the hydropower station, P max1 If the load is at its maximum for the day, it can perfectly handle peak shaving tasks; otherwise, it may cause problems such as... t p3 and t p4 The peak-shaving power shortage in the region. Figure 1 The blue area represents the base load diagram. If the capacity of the unit under maintenance is too large, it can cause... t p2 Base load deficit over a given period. Hydropower stations typically have a minimum output value. If the minimum output value is higher than the real-time load demand, it will cause a base load deviation, leading to power curtailment. Figure 1 The report did not provide hydrological data for the flood season. However, based on historical patterns of run-of-river hydropower stations, if hydropower unit maintenance is scheduled during the flood season, some water that would otherwise be used for power generation will be forced to be directly discharged due to unit shutdown, leading to an increase in water wastage. In summary, coordinating and optimizing the medium- and long-term discharge and maintenance plans of hydropower stations can significantly reduce the penalty costs associated with water wastage, base load deficits, and peak load deficits, and greatly improve system stability.

[0051] based on Figure 2 The improved ant colony algorithm solution steps are given.

[0052] 1) Initialization. Set the total number of iterations in the algorithm. N C The activity control parameters for each ant are randomly constructed, and the ants are randomly placed on various virtual nodes.

[0053] Let the number of ants be...m， ant colony size n , p Here, α is the pheromone volatile factor, and α is the pheromone factor. Q Where is the pheromone constant. tabu k For the first k A taboo list for each ant, used to store the set of virtual nodes that have already been visited. Node i and nodes j The pheromone concentration on the virtual edge between them is .

[0054] 2) Individual ants k From virtual node i To virtual node j random transition probability As shown in equation (28), Represents all virtual nodes that satisfy the constraints. j A set of.

[0055] Next, the individual ant selects the next virtual node and adjusts its tabu list until all nodes have been visited. The objective function of the path taken by different individual ants is calculated, as shown in equation (28).

[0056] (28) 3) Each ant locally updates the pheromone concentration according to Equation 5.

[0057] 4) The pheromone concentration of the entire population is updated according to Equation 6.

[0058] 5) Ant colony evolution strategy. Iteration. NE After that, the fitness of each ant is calculated according to Equation 29. Elite ants are recorded and saved and directly transferred to the offspring, while ordinary ants are selected, crossovered, and mutated to obtain the offspring ant colony.

[0059] (29) The global pheromone attenuation coefficient is adaptively adjusted according to equation (8), where for Pheromones attenuation coefficient, and These are the upper and lower limits of pheromone attenuation. The optimal value for the current iteration objective. This represents the optimal value for the objective across all historical iterations.

[0060] 7) If the algorithm still does not reach the termination condition, clear the tabu list of each ant and switch to step 2); otherwise, switch to step 8.

[0061] 8) Output the optimal solution.

[0062] Example 5: This invention proposes a joint optimization method based on an improved ant colony algorithm. The core of this method lies in solving the aforementioned background technical problems. The innovations include a multi-objective optimization framework for peak shaving: constructing a comprehensive objective function (Equation 1), and normalizing the weighted water abandonment penalty (…). C spin =0.28 yuan / kW The algorithm incorporates peak load deficit penalties (Cpeak = 0.55 yuan / kW) and base load deficit penalties (Cbase = 1.25 yuan / kW) to achieve a quantitative trade-off among the three objectives. An improved ant colony algorithm is employed: a multi-population co-evolution mechanism is used, dividing the ant colony into ordinary ant colonies (90%) and elite ant colonies (10%). Ordinary ant colonies utilize a Gaussian function (Equation 2-4) to enhance local search diversity; elite ant colonies optimize convergence speed based on probabilistic selection. A pheromone diffusion mechanism (local updates as shown in Equation 5 and global updates as shown in Equation 6) is introduced to improve global search capabilities and avoid premature convergence.

[0063] Dual-population collaborative architecture: one group optimizes maintenance timing (virtual node definition), and the other group optimizes unit combination strategy (operation mode), and handles high-dimensional solution space through co-evolution.

[0064] Mathematical model support: Define a complete set of constraints (Equations 9-27), including maintenance continuity, water balance and unit output constraints, to ensure the feasibility of the scheme.

[0065] Beneficial effects: Case study (a 3190W installed capacity hydropower station) shows that this invention can reduce water wastage losses and ensure peak-shaving capacity. Figure 5 ), Eliminating base load deficits ( Figure 4 The total operating cost is reduced by 15.4%-24.1% (Table 1), which is significantly better than the traditional algorithm.

[0066] Table 1. Comparison of annual costs under the three options

[0067] Example 6: The following describes the implementation of the present invention in detail with examples and case studies: 1. Data preparation: Hydropower station parameters: 21 generating units (total installed capacity 3190W), including 2 Class A units (170W / unit) and 19 Class B units (150W / unit); 12 units are awaiting maintenance (2 Class A units + 10 Class B units), with a maintenance duration of 14 days. The maximum number of units under maintenance concurrently is 2. Input data: historical power generation curves, inflow rate (…). Figure 6 ), power grid peak-shaving demand ( Figure 5 ), base load power ( Figure 4Algorithm parameter settings: ant colony size m = 50, pheromone factor α = 1, heuristic function factor β = 2, maximum number of iterations 9max = 200, global pheromone decay upper and lower limits 7in = 0.1, 7max = 0.9.

[0068] 2. The optimization process implementation steps are as follows: Step 1: Initialize the model; Construct objective functions and constraints in MATLAB 2024a.

[0069] Step 2: Run the improved ant colony algorithm; Initialize the ant colony and randomly place them on virtual nodes.

[0070] Ant path selection: Select the next virtual node based on the transition probability formula (maintenance decision). Pheromone update: Perform local and global updates, adaptively adjusting the decay coefficient; Evolutionary strategy: After each generation iteration, elite ant colonies are directly retained, while ordinary ant colonies generate new solutions through a Gaussian function.

[0071] Step 3: Output the optimization results; Generate annual maintenance plan ( Figure 3 The maintenance schedule for Category A units is arranged in January, while that for Category B units is concentrated in April-June; the power generation plan ensures base load matching. Figure 4 ) and peak-shaving capability ( Figure 5 ).

[0072] 3. Effect Verification: Results Comparison: Compared with the traditional ant colony algorithm and the control group particle swarm algorithm ( Figures 7-8 The total penalty cost of the present invention is reduced to RMB 93.4247 million (Table 1), a reduction of 15.4%-24.1%.

[0073] Key indicators: Water release penalty of RMB 18.4732 million, peak load deficit penalty of RMB 74.9515 million, base load deficit penalty of RMB 0: Maintenance period avoids peak hydrological period (July-August) to reduce water release; maintenance during normal water period utilizes reservoir capacity regulation to avoid resource waste.

[0074] Application expansion: This method can be integrated into the intelligent decision-making system of hydropower stations to support wind-solar hybrid dispatch scenarios. In the future, it can be combined with equipment health assessment to optimize the flexible constraints of maintenance time.

[0075] 4. Hardware and software environment: Hardware: Intel Core i7-13700KF CPU, 16GB RAM.

[0076] Software: The algorithm and simulation were implemented using the MATLAB 2024a platform.

[0077] This embodiment demonstrates the feasibility and superiority of the present invention, and is applicable to the annual planning optimization of large run-of-river hydropower stations.

[0078] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for joint optimization of annual maintenance and power generation plans for run-of-river hydropower stations based on an improved ant colony algorithm, characterized in that, The method includes: Step 1: Construct a multi-objective optimization model; A comprehensive objective function is constructed with the goal of minimizing annual water wastage losses, peak load deficits, and base load deficits. Step 2: Based on the actual power generation curves of the hydropower station over the years, the frequency analysis method is used to decompose the base load component and peak load. Step 3, define virtual nodes and virtual edges, virtual nodes U m,t Indicates the unit m At the initial time t Initial maintenance decision, virtual edge e m1 , e m2 Indicates from node U m1,t1 arrive U m2,t2 The transfer operation; Step 4: Optimize the maintenance period and operation mode using an improved multi-population co-evolutionary ant colony algorithm; the optimization process includes initialization, ant path selection, and local and global pheromone updates. Step 5: Output the optimal maintenance plan and power generation scheme to ensure that the annual water wastage loss, peak load deficit and base load deficit are minimized.

2. The method for joint optimization of annual maintenance and power generation plans for run-of-river hydropower stations based on an improved ant colony algorithm as described in claim 1, characterized in that, The objective function constructed in step 1 is: ；（1） In the formula: d For date, The amount of electricity converted from the daily water wastage. This is the penalty coefficient for water abandonment. This refers to the daily peak-shaving shortfall in electricity volume. This is the peak-shaving penalty coefficient. This refers to the daily baseload deficit. This is the baseload shortfall penalty coefficient.

3. The method for joint optimization of annual maintenance and power generation plans for run-of-river hydropower stations based on an improved ant colony algorithm as described in claim 2, characterized in that, In step 1, during the construction of the objective function, the three objectives are quantitatively balanced through normalized weighted water wastage penalty, peak load deficit penalty, and baseload deficit penalty.

4. The method for joint optimization of annual maintenance and power generation plans for run-of-river hydropower stations based on an improved ant colony algorithm as described in claim 1, characterized in that, The improved multi-population co-evolutionary ant colony algorithm in step 4 includes the following steps: Step 4.1: Divide the ant colony into ordinary ant colonies and elite ant colonies; Step 4.2: The ordinary ant colony generates a new solution based on the Gaussian function: Description of the Gaussian function for ordinary ants: ；（2） ；（3） ；（4） In the formula, It is the first Gaussian function of a common ant To optimize the solution to the problem, It is the first The generation The mean of a normal distribution, It is the first The generation The standard deviation of a normal distribution It is the first The generation The variance of a normal distribution, It is the first In the distribution, the th... The fitness value of each sample It is the first Scaling factor of the generation, Population size; Step 4.3: The elite ant colony uses a probabilistic selection method combined with a Gaussian kernel function to generate a solution; Step 4.4 introduces a pheromone diffusion mechanism to improve global search capabilities and avoid premature convergence; Step 4.5: Adaptively adjust the global pheromone attenuation coefficient.

5. The method for joint optimization of annual maintenance and power generation plans for run-of-river hydropower stations based on an improved ant colony algorithm as described in claim 4, characterized in that, In step 4.1, elite ant colonies account for 10%, while ordinary ant colonies account for 90%.

6. The method for joint optimization of annual maintenance and power generation plans for run-of-river hydropower stations based on an improved ant colony algorithm as described in claim 5, characterized in that, Step 4.4 introduces the pheromone diffusion mechanism, specifically including: The local pheromone update formula is as shown in equation (5), which updates the maintenance node. To the maintenance node Pheromones between them: ；（5） In the formula, Let be the number of iterations of the ant colony. Indicates the rate at which an individual's pheromones evaporate. The range of values ​​is ∈ (0,1), Indicates the local pheromone residue coefficient. For ants from the maintenance node To the maintenance node The increase in pheromones between them For the current iteration N At this time, node to node The concentration of pheromones along the path between them After the update The next iteration; The global pheromone update formulas are as shown in equations (6) and (7): ；（6） ；（7） According to equations (6) and (7), This indicates after the update, i.e. Global pheromone levels after the next iteration; It is the present moment, that is Pheromones concentration after the next iteration It is the global pheromone evaporation rate; It is the global pheromone growth coefficient. The current objective function value, if It was before The minimum value in the next iteration, i.e., the optimal value, is the pheromone increment. for If not, then It is 0.

7. The method for joint optimization of annual maintenance and power generation plans for run-of-river hydropower stations based on an improved ant colony algorithm as described in claim 6, characterized in that, In step 4.5, the global pheromone attenuation coefficient is adaptively adjusted as shown in equation (8): ；（8） In the formula, for Pheromones attenuation coefficient, and These are the upper and lower limits of pheromone attenuation. The optimal value for the current iteration objective. This represents the optimal value for the objective across all historical iterations.

8. The method for joint optimization of annual maintenance and power generation plans for run-of-river hydropower stations based on an improved ant colony algorithm as described in claim 1, characterized in that, It also includes constraint handling steps: 1) Continuity constraints for unit maintenance, as shown in equation (9): ； （9） In the formula, For unit maintenance status variables, when the unit In the d During daily maintenance, its value is 1; otherwise, its value is 0. For the unit Maintenance start date. For the unit Maintenance duration; For hydropower unit maintenance, the unit maintenance must be completed continuously in one go to avoid interruption leading to reduced maintenance efficiency or equipment damage risk. 2) Constraints for synchronous maintenance of generating units, as shown in formula (10): ；（10） In the formula, A collection of all units awaiting maintenance. This is the maximum number of generating units that can be maintained during the same period; run-of-river hydropower stations need to maintain enough operating units to meet basic power generation and peak load requirements; if multiple units are under maintenance at the same time, it may lead to insufficient power output of the power station, exacerbating water wastage or peak load / base load power shortages. 3) Unit start-up and shutdown constraints, as shown in equations (11)-(13): ；（11） ；（12） ；（13） In equations (11)-(13), For the unit m In the d Daytime t The start / stop status variable has a value of 1 when it is in the power-on state and a value of 0 otherwise. , These represent the duration of continuous unit startup / shutdown. 4) Water balance constraint, as shown in equation (14): ； （14） As shown in equation (14). and The respective d sky t Time period and t- Storage capacity for one time period; This represents the time step; the constant 3600 represents 1. h Equals 3600s. and The first d Daytime t The inflow and outflow of the hydropower station m 3 / s ; 5) Outbound flow constraints, as shown in equations (15) and (16): ；（15） ；（16） As shown in equations (15)-(16), t Outbound flow during different time periods Power generation flow of the power plant and water discharge flow sum; and These are the maximum and minimum outbound flow limits, respectively. 6) Output constraints of hydropower units, as shown in equations (17) and (18): ；（17） ；（18） In the formula, NHQ represents the dynamic characteristics of the unit, and represents the nonlinear functional relationship between the unit's output and the water head and power generation flow rate. The generating flow rate of the unit is m 3 / s, For the generator head of the unit, m ; , For the unit m Maximum / minimum technical output limits; 7) Gradient constraint of hydropower unit, as shown in equation (19): ； （19） In the formula, and Hydropower station units m Time period t Internal output limits for downward and upward climbing; 8) Vibration zone constraint, as shown in equation (20): ； （20） In the formula, , For the unit m The k The upper and lower limits of output in each vibration zone; 9) Generating head constraint of the unit, as shown in equation (21): ； （21） In the formula, For the unit m In the d Daytime t Hydropower head; For the first d Daytime t The reservoir water level; For the first d Daytime t The tailwater level; 10) Reservoir water level constraints, as shown in equations (22) and (23): ；（22） ； （23） In the formula, For the reservoir water level, and The first d Minimum / maximum water level limits for the day This is the maximum allowable fluctuation in water level for that day. 11) Water level-reservoir capacity constraint, as shown in equation (24): ； （24） In the formula, This is the water level-reservoir capacity curve function of the power station. For storage capacity; 12) Tailwater level - outflow constraint, as shown in equation (25): ；（25） In the formula, This is a function of the tailrace level and outflow rate of the power station. 13) Upper and lower peak-shaving capacity constraints: ；（26） ；（27） In the formula, For the unit m In the d Heavenly t Peak capacity adjustment during specific time periods Ramp-up restrictions for unit peak shaving For the unit m In the d Heavenly t The operational status of a time period is represented by a binary variable, where 1 indicates operation and 0 indicates shutdown. For the unit m The upper limit of output, For the unit m In the d Heavenly t Actual power generation during the period Non-negative constraints for peak capacity adjustment; For the unit m In the d Heavenly t Peak-shaving capacity during the period, Ramp-up restrictions for unit peak shaving For the unit m The lower limit of output, Non-negative constraint for peak reduction capacity; As shown in equations (26) and (27), the operating generating units reserve up and down adjustment capabilities to meet the peak shaving needs of the power grid.

9. A joint optimization system for annual maintenance and power generation plans of run-of-river hydropower stations based on an improved ant colony algorithm, characterized in that, The system is used to implement the joint optimization method for annual maintenance and power generation plans of run-of-river hydropower stations based on an improved ant colony algorithm as described in any one of claims 1-8, including: Data input module: used to import historical power generation curves of hydropower stations, unit parameters, inflow data and grid peak shaving requirements; Optimization Engine Module: Performs joint optimization based on the improved ant colony algorithm, including collaborative search, pheromone update and diffusion between ordinary ant colonies and elite ant colonies; Simulation calculation module: Implements objective function calculation and constraint verification in the MATLAB environment; Output module: Generates annual maintenance plan and power generation operation mode diagram.

10. The joint optimization system for annual maintenance and power generation planning of run-of-river hydropower stations based on an improved ant colony algorithm as described in claim 9, characterized in that, Unit parameters include installed capacity and maintenance cycle.

Images