A scheduling method of a water, wind and light complementary power generation system based on a knowledge graph structure

Abstract

The application provides a kind of scheduling method of water, wind and light complementary power generation system based on knowledge graph structure belongs to mixed energy scheduling technical field.The application designs water, wind and light complementary power generation system scheduling rule structure based on "object-time period-situation-decision" knowledge graph, extracts the variable set that can be optimized in knowledge graph rule, establishes water, wind and light complementary power generation system knowledge graph scheduling rule optimization model and completes solution, proposes the graph chain reasoning method based on current scheduling situation to obtain decision result, obtains the scheduling scheme calculated according to knowledge graph rule of water, wind and light complementary power generation system.The knowledge graph scheduling rule of the application is compared with the rule of traditional scheduling function form, and the power generation capacity is higher, the satisfaction degree to power grid load is higher, meanwhile, the knowledge graph scheduling rule can process multi-objective scheduling demand, has the advantages of good intuitiveness, strong explainability, flexible and easy to expand, etc., and has important significance for operation and decision of water, wind and light complementary power generation system.

Description

Technical Field

[0001] This invention belongs to the field of hybrid energy system scheduling technology, and more specifically, relates to a scheduling method for a hydro-wind-solar hybrid power generation system based on a knowledge graph structure. Background Technology

[0002] Currently, the commonly used technologies in the dispatching industry for hydro-wind-solar hybrid power generation systems are as follows:

[0003] Hybrid energy system dispatch is a technology that considers the joint operation of multiple energy subsystems such as hydropower, wind power, or photovoltaic power generation. Hydro-wind-solar hybrid power generation systems are currently the most important typical hybrid system in hybrid energy system dispatch. These systems exhibit strong randomness, volatility, and uncertainty, increasing the difficulty of obtaining dispatch schemes. Therefore, dispatch rules for hydro-wind-solar hybrid power generation systems are needed to assist dispatch decision-makers in creating dispatch plans. These dispatch rules are of great significance to the operation and decision-making of hydro-wind-solar hybrid power generation systems.

[0004] Against the backdrop of dwindling non-renewable energy reserves and increasingly severe environmental damage, developing clean and renewable energy sources such as hydropower, wind power, and solar power, and researching complementary scheduling strategies among them, has become an important direction for my country's water resources and energy sector. Wind and solar resources are spatially unevenly distributed and temporally unstable, exhibiting strong randomness, volatility, and intermittency. This makes it difficult for independent wind and solar subsystems to continuously output stable power, restricting the grid's capacity to absorb wind and solar energy. Hydropower's rapid regulation capabilities and strong storage capacity can effectively mitigate the adverse effects of intermittent energy output fluctuations on the power system. Furthermore, hydropower output is complementary to wind and solar power output at different time scales within the year and day. Utilizing the strong regulation advantages of hydropower and its complementary characteristics with wind and solar power output, packaging and jointly operating the outputs of the three subsystems can effectively solve the problem of grid integration for centralized wind and solar power. The hydro-wind-solar multi-energy complementary system is simultaneously affected by multiple uncertainties in its three subsystems. Uncertainties in hydro-wind-solar forecasts introduce cascading uncertainties into the joint scheduling of the complementary system, and scheduling uncertainties, in turn, introduce cascading uncertainties into operational decisions, exacerbating the instability risks associated with mitigating power output fluctuations. Assisting decision-making through scheduling rules in uncertain scheduling scenarios is a practical problem that urgently needs to be addressed for the joint operation of the hydro-wind-solar multi-energy complementary system under uncertain conditions.

[0005] The problem with existing technology is:

[0006] The deterministic optimization scheduling of hydro-wind-solar multi-energy complementary systems faces risks in guiding real-time operation due to the lack of consideration for uncertainties in forecasts such as runoff, wind speed, and solar radiation intensity. Research on uncertain scheduling rules typically employs two approaches: rule extraction and rule optimization. The former involves excessive sample preparation time and suffers from dimensionality issues when extracting multi-objective scheduling rules. Rule optimization studies employ simplistic scheduling graph formats, limiting the rules they can represent and making them less operable for rules under multi-factor combinations in hydro-wind-solar complementary systems. Rules in the form of scheduling functions are limited in their application scenarios due to their poor intuitiveness. Therefore, designing an intuitive and adaptable scheduling rule format that can optimize multi-objective scheduling rules for hydro-wind-solar complementary systems under uncertain scheduling scenarios is a pressing issue that needs to be addressed. Summary of the Invention

[0007] To address the aforementioned deficiencies or improvement needs of existing technologies, this invention provides a scheduling method for a hydro-wind-solar hybrid power generation system based on a knowledge graph structure. By employing this invention, intuitive and highly adaptable multi-objective scheduling rules for hydro-wind-solar hybrid power generation systems can be obtained.

[0008] To achieve the aforementioned technical features, the present invention aims to provide a scheduling method for a hydro-wind-solar hybrid power generation system based on a knowledge graph structure. This method first designs a scheduling rule structure for the hydro-wind-solar hybrid power generation system based on an "object-time period-situation-decision" knowledge graph. Then, it extracts the set of variables that can be optimized from the knowledge graph rules, establishes an optimization model for the knowledge graph scheduling rules of the hydro-wind-solar hybrid power generation system, and solves the model. Finally, it uses a graph chain reasoning method based on the current scheduling situation to obtain the decision result and acquire the scheduling scheme calculated by the hydro-wind-solar hybrid power generation system according to the knowledge graph rules.

[0009] Preferably, the scheduling method specifically includes the following steps:

[0010] S1 collects basic attribute data, scheduling characteristic curve data, hydrological data, and power output data of each power station in the hydro-wind-solar hybrid power generation system;

[0011] S2, based on the knowledge graph of "object-time period-situation-decision", design the scheduling rule structure of the hydro-wind-solar hybrid power generation system, and clarify the entities, attributes and relationships in the knowledge graph scheduling rules;

[0012] S3, extract the set of variables that can be optimized from the knowledge graph scheduling rules, and use them as decision variables for the scheduling rule optimization model;

[0013] S4 aims to maximize power generation and achieve the highest degree of power output matching with load. It constructs constraints for three subsystems: hydropower, wind power, and photovoltaic power generation. It establishes a knowledge graph scheduling rule optimization model for the hydro-wind-solar hybrid power generation system and solves the problem based on a multi-objective evolutionary algorithm.

[0014] S5. After the knowledge graph scheduling rules of the hydro-wind-solar hybrid power generation system are optimized, a graph chain reasoning method is proposed to obtain the decision result based on the current scheduling situation.

[0015] S6, based on knowledge graph rules and graph chain reasoning method, obtains the scheduling scheme of the hydro-wind-solar hybrid power generation system calculated according to knowledge graph rules.

[0016] Preferably, the knowledge graph scheduling rule structure of "object-time period-situation-decision" in S2 is specifically as follows:

[0017] (1) The scheduling object graph establishes the relationships between hydropower, wind power, and photovoltaic computing entities through hydraulic topology connections, multi-site proximity relationships, and hydro-wind-solar complementarity relationships, and assigns characteristic values, characteristic curves, and scheduling constraint attributes to each computing entity; the hydro-wind-solar complementary power generation system has M respectively w M s M h If there are wind power stations, photovoltaic power stations, and hydropower stations, then M will be included in the dispatch object map. w M s M h One wind power station entity, one photovoltaic power station entity, and one hydropower station entity;

[0018] (2) The scheduling time period map mainly establishes the time period connection relationship for each scheduling time period entity, and clarifies the time period index, time period start time, time period end time, time scale, and total number of scheduling time periods; the entire scheduling calculation period has T time periods, then from 0 to (T-1) are all scheduling time period map entities, where time period 0 is the start time period entity, time period (T-1) is the end time period entity, and time period t and time period t+1 are "immediately connected";

[0019] (3) The scheduling situation map establishes the situation entity for the facing time period from three aspects: meteorology, water conditions and engineering conditions. The inflow, water level on the dam, wind speed or wind power output, solar radiation intensity or photovoltaic output for the facing time period are used as situation attributes. The combination relationship of each scheduling situation and the closed loop relationship formed by all situations are considered.

[0020] The dispatch situation map has multiple condition types: inflow Qi, wind and solar power output Nn, wind power output Nw, solar power output Ns, and upstream water level Zu. The dispatch situation map is one or more of the above condition types. M types of conditions are taken as the dispatch situation map, and each type of condition is divided into C. MIf there are several condition segments, then the scheduling situation has... Given the following scenario combinations, assuming the first condition type is the inbound traffic type Qi, within its possible value range... If the middle is divided into segment C1, then the first set of condition nodes is: There are a total of C1 scheduling situation nodes, and M types of conditions. There are several scheduling situation nodes. The first group of scheduling situation nodes and time period nodes are "linked" to each other. Nodes of the same type are "closed-loop" to each other, and nodes of different types are "combined" to each other.

[0021] (4) The scheduling decision map completes the scheduling decision for water, wind, and solar power during the specified time period based on the scheduling object map, scheduling time period map, and scheduling situation map. It uses various decision results as entities, the water level at the end of the time period, the reservoir outflow during the time period, or the hydropower output decision value during the time period as attributes, and the relationship between the situation conditions and the decision results as the relation. The scheduling decision map also has multiple types: outflow Qo, upstream water level Zu. Only one type can be used as the decision variable in the scheduling decision map. Each combination of scheduling situations corresponds to a scheduling decision node, M. c There are a total of M possible combinations of situations. c If the outbound flow is used as the scheduling decision variable type for a scheduling decision node, then the scheduling decision node is...

[0022] Preferably, in step S3, extracting the set of variables that can be optimized from the knowledge graph scheduling rules specifically involves:

[0023] The scheduling object map and the scheduling time period map can be established after the scheduling requirements are determined, and they are deterministic, meaning there are no variables that need to be optimized.

[0024] After the scheduling demand is determined, only the structure of the scheduling situation graph and scheduling decision graph can be determined. However, the specific condition values ​​for each scheduling situation entity and the decision values ​​for each scheduling decision entity are still uncertain and need to be optimized. For the t-th time period, assuming M types of conditions are used as the scheduling situation graph, each type of condition can be divided into C... M Given a condition segment with a beginning and an end, C will be generated. M +1 scheduling situation variable to be optimized, M types of conditions will generate a total of There are several scheduling situation variables to be optimized, namely: Type M conditions will occur There are several scheduling situation combinations, and each scheduling situation combination corresponds to a scheduling decision variable, i.e., there are... One scheduling decision variable to be optimized: In the t-th time period, a total of There are 12 variables to be optimized, which will be generated over a total of T scheduling periods. There are 1 variable to be optimized.

[0025] Preferably, in step S4, the objective, constraints, and solution method of the knowledge graph scheduling rule optimization model for the hydro-wind-solar hybrid power generation system are as follows:

[0026] (1) Scheduling objective:

[0027] 1) Objective 1: Maximize power generation:

[0028]

[0029] In the formula, E(N) is the total power generation of the hydro-wind-solar hybrid power generation system; N t This represents the power output of the hydro-wind-solar hybrid power generation system during time period t; T and Δt represent the total number of scheduling periods and the duration of a single scheduling period; N w,i,t N s,j,t and N h,k,t M represents the power output of the i-th wind power station, the j-th photovoltaic power station, and the k-th hydropower station in the t-th time period; w M s M h This represents the number of wind power stations, photovoltaic power stations, and hydropower stations in a hydro-wind-solar hybrid power generation system.

[0030] 2) Objective 2: Achieving the highest degree of output-load matching.

[0031]

[0032] In the formula, It refers to the degree of matching between output and load; P t This represents the power load during the t-th time period;

[0033] (2) Scheduling constraints:

[0034] 1) Reservoir water level constraints:

[0035]

[0036] In the formula, Zu i,t It is the upstream water level of the i-th reservoir in the t-th time period. and These are its upper and lower limits under flood control constraints;

[0037] 2) Reservoir discharge constraints:

[0038]

[0039] In the formula, Qo i,tIt is the outflow from the i-th reservoir in the t-th time period. and These are its corresponding upper and lower limit constraints; outbound flow Qo i,t Power generation flow rate Qg i,t and the discharge flow rate Qa i,t Composition, power generation flow rate Qg i,t Cannot exceed the maximum transmission capacity

[0040] 3) Hydropower output constraints:

[0041]

[0042] In the formula, Nh i,t It represents the output of the i-th reservoir in the t-th time period. and These are its corresponding upper and lower limit constraints; hydropower output is also constrained by the expected output curve and the NHQ curve.

[0043] 4) Water balance equation:

[0044] V i,t+1 =V i,t +(Qi i,t -Qo i,t -Ev i,t )·Δt;(8)

[0045] In the formula: V i,t and V i,t+1 Qi represents the reservoir capacity of the i-th reservoir at the beginning and end of the t-th time period, respectively. i,t and Qo i,t These are the corresponding inbound and outbound traffic, Ev i,t This represents water loss due to evaporation;

[0046] 5) Hydraulic connections between cascade hydropower stations:

[0047]

[0048] In the formula, Qitv i,t It is the interval flow of the river section upstream of the i-th reservoir; It is the traffic of upstream sites. The flow rate when the river reaches the i-th reservoir through its course evolution; Φ i It is the set of upstream stations that are hydraulically connected to the i-th reservoir. It is Φ i Element; This represents a river channel evolution model, specifically the Muskingen model or the time-delay evolution model.

[0049] 6) Wind power constraints:

[0050]

[0051] In the formula, v j,t It is the wind speed of the j-th wind power station in the t-th time period. and It refers to the inlet and outlet wind speeds; Nw j,t It represents the power output of the j-th wind power station in the t-th time period, and it cannot exceed the installed capacity of the corresponding wind power station.

[0052] 7) Constraints on photovoltaic power generation:

[0053]

[0054] In the formula, Ns k,t and These represent the power output and installed capacity of the k-th photovoltaic power station, respectively.

[0055] 8) Conveying channel capacity constraints:

[0056]

[0057] In the formula, N g,t and These represent the packaged output and transmission channel capacity of the g-th group of water-wind-solar multi-energy complementary systems, respectively.

[0058] (3) Solution method:

[0059] The optimization model for the knowledge graph scheduling rules of the hydro-wind-solar hybrid power generation system is a multi-objective optimization problem. Based on this, an arbitrary multi-objective evolutionary algorithm is used to solve it.

[0060] Preferably, in step S5, the graph chain reasoning method for obtaining the decision result based on the current scheduling situation specifically includes:

[0061] Chain reasoning is completed in five steps:

[0062] S5.1, Based on the scheduling object map, select the hydropower, wind power, and photovoltaic objects participating in this scheduling calculation, and determine the complementary scheduling relationship;

[0063] S5.2, Select the scheduling period t from the scheduling period map;

[0064] S5.3 First, based on the wind power and solar power dispatch map objects in step S5.1 and the time period calculation reasoning in step S5.2, the sum of wind power and solar power output for the time period is obtained. Then, based on the hydropower dispatch map objects in step S5.1 and the time period for the time period, the initial water level above the dam for the time period is obtained.

[0065] S5.4, Based on the reasoning of the two-dimensional variables in the scheduling situation graph in step S5.3, the time period hydropower output decision value under this scheduling situation combination is obtained based on the scheduling decision graph reasoning.

[0066] S5.5 Assign the reasoning result of step S5.4 to the hydropower scheduling object, and obtain the scheduling process of the reservoir upstream reservoir and reservoir outflow at the end of the reservoir period based on the water balance equation and hydropower output calculation formula.

[0067] The five steps S5.1 to S5.5 are repeated from the start of the scheduling period to the end of the scheduling period to complete the scheduling decision according to the knowledge graph rules of the water-wind-solar hybrid system throughout the entire scheduling period.

[0068] Preferably, another aspect of the present invention provides a computer program for a knowledge graph-based scheduling method for a hydro-wind-solar hybrid power generation system, wherein the computer program is used to implement the knowledge graph-based scheduling method for the hydro-wind-solar hybrid power generation system.

[0069] Preferably, another aspect of the present invention provides a terminal, the terminal being equipped with at least a controller that implements the knowledge graph-based hydro-wind-solar hybrid power generation system scheduling method.

[0070] Preferably, another aspect of the present invention provides a computer-readable storage medium including instructions that, when executed on a computer, cause the computer to perform the aforementioned knowledge graph-based hydro-wind-solar hybrid power generation system scheduling method.

[0071] Preferably, another aspect of the present invention provides a control system for a knowledge graph-based scheduling method for a hydro-wind-solar hybrid power generation system, the control system being used to implement the knowledge graph-based scheduling method for the hydro-wind-solar hybrid power generation system described in the claims.

[0072] Preferably, another aspect of the present invention provides a hydro-wind-solar hybrid power generation system scheduling device equipped with the control system of the knowledge graph-based hydro-wind-solar hybrid power generation system scheduling method.

[0073] The present invention has the following beneficial effects:

[0074] 1. This invention uses a knowledge graph rule structure to optimize the scheduling rules of a hydro-wind-solar hybrid power generation system. This rule does not restrict the form of the rule and has the advantages of being intuitive and clear. It is highly practical and applicable, and can be used to obtain multi-objective scheduling rules. It can guide dispatchers to make scheduling plans for hydro-wind-solar hybrid power generation systems.

[0075] 2. This invention uses a knowledge graph structure as the form of scheduling rules, which has strong scalability. When the scheduling object changes or the scheduling period is extended, it is only necessary to expand the scheduling rules corresponding to the changed scheduling object and the increased scheduling period. The previously optimized scheduling rules can be retained and used, without completely changing the scheduling rules of the entire object and the entire period, thus saving the adjustment costs caused by changes in scheduling requirements. Attached Figure Description

[0076] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0077] Figure 1 This is a flowchart of scheduling rules for a hydro-wind-solar hybrid power generation system based on a knowledge graph, provided by an embodiment of the present invention.

[0078] Figure 2 This is a schematic diagram of the knowledge graph scheduling rules provided in an embodiment of the present invention.

[0079] Figure 3 This is a schematic diagram of the variables to be optimized provided in an embodiment of the present invention.

[0080] Figure 4 A schematic diagram of the knowledge graph scheduling rule chain reasoning method provided in this embodiment of the invention.

[0081] Figure 5 The multi-objective scheduling rule provided in this embodiment of the invention has a non-inferiority frontier.

[0082] Figure 6 The target value provided in this embodiment of the invention is compared with the traditional scheduling function form rule.

[0083] Figure 7 The present invention provides a scheduling process calculated by scheduling rules.

[0084] Figure 8 The present invention provides an output stacking diagram calculated by scheduling rules.

[0085] Figure 9 The knowledge graph scheduling rules provided in this embodiment of the invention. Detailed Implementation

[0086] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments, examples of which are shown in the drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0087] Example 1:

[0088] Please see Figure 1-9 This invention designs a scheduling rule structure for a hydro-wind-solar hybrid power generation system based on a knowledge graph of "object-time period-situation-decision". It extracts the set of variables that can be optimized from the knowledge graph rules, establishes and solves the knowledge graph scheduling rule optimization model for the hydro-wind-solar hybrid power generation system, and proposes a graph chain reasoning method to obtain decision results based on the current scheduling situation, thereby obtaining the scheduling scheme calculated by the hydro-wind-solar hybrid power generation system according to the knowledge graph rules.

[0089] Appendix Figure 1 The diagram shown is a flowchart of the scheduling rules for a knowledge graph-based hydro-wind-solar hybrid power generation system, which includes the following steps:

[0090] S1 collects basic attribute data, scheduling characteristic curve data, hydrological data, and power output data of each power station in the hydro-wind-solar hybrid power generation system;

[0091] S2. Based on the "object-time period-situation-decision" knowledge graph, design the scheduling rule structure for the hydro-wind-solar hybrid power generation system, and clarify the entities, attributes, and relationships in the knowledge graph scheduling rules; a schematic diagram of the knowledge graph scheduling rules is attached. Figure 2 As shown.

[0092] (1) The scheduling object graph establishes the relationships between hydropower, wind power, and photovoltaic computing entities through hydraulic topology connections, multi-site proximity relationships, and hydro-wind-solar complementarity relationships, and assigns characteristic values, characteristic curves, and scheduling constraint attributes to each computing entity; the hydro-wind-solar complementary power generation system has M respectively w M s M h If there are wind power stations, photovoltaic power stations, and hydropower stations, then M will be included in the dispatch object map. w M s M h One wind power station entity, one photovoltaic power station entity, and one hydropower station entity;

[0093] (2) The scheduling time period map mainly establishes the time period connection relationship for each scheduling time period entity, and clarifies the time period index, time period start time, time period end time, time scale, and total number of scheduling time periods; the entire scheduling calculation period has T time periods, then from 0 to (T-1) are all scheduling time period map entities, where time period 0 is the start time period entity, time period (T-1) is the end time period entity, and time period t and time period t+1 are "immediately connected";

[0094] (3) The scheduling situation map establishes the situation entity for the facing period from three aspects: meteorology, water conditions and engineering conditions. The inflow, water level on the dam, wind speed (or wind power output) and solar radiation intensity (or photovoltaic power output) of the facing period are used as situation attributes. The combination relationship of each scheduling situation and the closed loop relationship formed by all situations are considered (without omissions or redundancy).

[0095] The dispatch situation map has multiple condition types: inflow Qi, wind and solar power output Nn, wind power output Nw, solar power output Ns, and upstream water level Zu. The dispatch situation map is one or more of the above condition types. M types of conditions are taken as the dispatch situation map, and each type of condition is divided into C. M If there are several condition segments, then the scheduling situation has... Given the following scenario combinations, assuming the first condition type is the inbound traffic type Qi, within its possible value range... If the middle is divided into segment C1, then the first set of condition nodes is: There are a total of C1 scheduling situation nodes, and M types of conditions. There are several scheduling situation nodes. The first group of scheduling situation nodes and time period nodes are "linked" to each other. Nodes of the same type are "closed-loop" to each other, and nodes of different types are "combined" to each other.

[0096] (4) The scheduling decision map completes the scheduling decision for water, wind, and solar power during the specified time period based on the scheduling object map, scheduling time period map, and scheduling situation map. It uses various decision results as entities, the water level at the end of the time period, the reservoir outflow during the time period, or the hydropower output decision value during the time period as attributes, and the relationship between the situation conditions and the decision results as the relation. The scheduling decision map also has multiple types: outflow Qo, upstream water level Zu, etc. However, the scheduling decision map can only take one type as the decision variable. Each combination of scheduling situations corresponds to a scheduling decision node, M. c There are a total of M possible combinations of situations. c If the outbound flow is used as the scheduling decision variable type for a scheduling decision node, then the scheduling decision node is...

[0097] The detailed rules of the "Object-Time Period-Situation-Decision" knowledge graph, covering entities, attributes, and relationships, are shown in the table below.

[0098] Table 1: Explanation of the rule structure of the "Object-Time Period-Situation-Decision" knowledge graph

[0099]

[0100] S3 extracts a set of optimizable variables from the knowledge graph scheduling rules, which serve as the decision variables for the scheduling rule optimization model, as shown in the appendix. Figure 3 As shown.

[0101] The scheduling object map and the scheduling time period map can be established after the scheduling requirements are determined, and they are deterministic, meaning there are no variables that need to be optimized.

[0102] After the scheduling demand is determined, only the structure of the scheduling situation graph and scheduling decision graph can be determined. However, the specific condition values ​​for each scheduling situation entity and the decision values ​​for each scheduling decision entity are still uncertain and need to be optimized. For the t-th time period, assuming M types of conditions are used as the scheduling situation graph, each type of condition can be divided into C... M Given a condition segment with a beginning and an end, C will be generated. M +1 scheduling situation variable to be optimized, M types of conditions will generate a total of There are several scheduling situation variables to be optimized, namely: Type M conditions will occur There are several scheduling situation combinations, and each scheduling situation combination corresponds to a scheduling decision variable, i.e., there are... One scheduling decision variable to be optimized: In the t-th time period, a total of There are 12 variables to be optimized, which will be generated over a total of T scheduling periods. There are 1 variable to be optimized.

[0103] S4 aims to maximize power generation and achieve the highest degree of power output matching with load. It constructs constraints for three subsystems: hydropower, wind power, and photovoltaic power generation. It establishes a knowledge graph scheduling rule optimization model for the hydro-wind-solar hybrid power generation system and solves the problem based on a multi-objective evolutionary algorithm.

[0104] (1) Scheduling objective:

[0105] 1) Objective 1: Maximize power generation:

[0106]

[0107] In the formula, E(N) is the total power generation of the hydro-wind-solar hybrid power generation system; N t This represents the power output of the hydro-wind-solar hybrid power generation system during time period t; T and Δt represent the total number of scheduling periods and the duration of a single scheduling period; N w,i,t N s,j,t and N h,k,t M represents the power output of the i-th wind power station, the j-th photovoltaic power station, and the k-th hydropower station in the t-th time period; w M s M h This represents the number of wind power stations, photovoltaic power stations, and hydropower stations in a hydro-wind-solar hybrid power generation system.

[0108] 2) Objective 2: Achieving the highest degree of output-load matching.

[0109]

[0110] In the formula, It refers to the degree of matching between output and load; P t This represents the power load during the t-th time period;

[0111] (2) Scheduling constraints:

[0112] 1) Reservoir water level constraints:

[0113]

[0114] In the formula, Zu i,t It is the upstream water level of the i-th reservoir in the t-th time period. and These are its upper and lower limits under flood control constraints;

[0115] 2) Reservoir discharge constraints:

[0116]

[0117] In the formula, Qo i,t It is the outflow from the i-th reservoir in the t-th time period. and These are its corresponding upper and lower limit constraints; outbound flow Qo i,t Power generation flow rate Qg i,t and the discharge flow rate Qa i,t Composition, power generation flow rate Qg i,t Cannot exceed the maximum transmission capacity

[0118] 3) Hydropower output constraints:

[0119]

[0120] In the formula, Nh i,t It represents the output of the i-th reservoir in the t-th time period. and These are its corresponding upper and lower limit constraints; hydropower output is also constrained by the expected output curve and the NHQ curve.

[0121] 4) Water balance equation:

[0122] V i,t+1 =V i,t +(Qi i,t -Qo i,t -Ev i,t )·Δt;(8)

[0123] In the formula: V i,t and V i,t+1 Qi represents the reservoir capacity of the i-th reservoir at the beginning and end of the t-th time period, respectively. i,t and Qo i,t These are the corresponding inbound and outbound traffic, Evi,t This represents water loss due to evaporation;

[0124] 5) Hydraulic connections between cascade hydropower stations:

[0125]

[0126] In the formula, Qitv i,t It is the interval flow of the river section upstream of the i-th reservoir; It is the traffic of upstream sites. The flow rate when the river reaches the i-th reservoir through its course evolution; Φ i It is the set of upstream stations that are hydraulically connected to the i-th reservoir. It is Φ i Element; This represents a river channel evolution model, specifically the Muskingen model or the time-delay evolution model.

[0127] 6) Wind power constraints:

[0128]

[0129] In the formula, v j,t It is the wind speed of the j-th wind power station in the t-th time period. and It refers to the inlet and outlet wind speeds; Nw j,t It represents the power output of the j-th wind power station in the t-th time period, and it cannot exceed the installed capacity of the corresponding wind power station.

[0130] 7) Constraints on photovoltaic power generation:

[0131]

[0132] In the formula, Ns k,t and These represent the power output and installed capacity of the k-th photovoltaic power station, respectively.

[0133] 8) Conveying channel capacity constraints:

[0134]

[0135] In the formula, N g,t and These represent the packaged output and transmission channel capacity of the g-th group of water-wind-solar multi-energy complementary systems, respectively.

[0136] (3) Solution method:

[0137] The optimization model for the knowledge graph scheduling rules of the hydro-wind-solar hybrid power generation system is a multi-objective optimization problem. Based on this, an arbitrary multi-objective evolutionary algorithm is used to solve it.

[0138] Preferably, in this embodiment, a multi-objective evolutionary algorithm is used to solve the problem.

[0139] S5. After the knowledge graph scheduling rules of the hydro-wind-solar hybrid power generation system are optimized, a graph chain reasoning method is proposed to obtain the decision result based on the current scheduling situation, as shown in the appendix. Figure 4 As shown.

[0140] Chain reasoning is completed in five steps:

[0141] S5.1, Based on the scheduling object map, select the hydropower, wind power, and photovoltaic objects participating in this scheduling calculation, and determine the complementary scheduling relationship;

[0142] S5.2, Select the scheduling period t from the scheduling period map;

[0143] S5.3 First, based on the wind power and solar power dispatch map objects in step S5.1 and the time period calculation reasoning in step S5.2, the sum of wind power and solar power output for the time period is obtained. Then, based on the hydropower dispatch map objects in step S5.1 and the time period for the time period, the initial water level above the dam for the time period is obtained.

[0144] S5.4, Based on the reasoning of the two-dimensional variables in the scheduling situation graph in step S5.3, the time period hydropower output decision value under this scheduling situation combination is obtained based on the scheduling decision graph reasoning.

[0145] S5.5 Assign the reasoning result of step S5.4 to the hydropower scheduling object, and obtain the scheduling process of the reservoir upstream reservoir and reservoir outflow at the end of the reservoir period based on the water balance equation and hydropower output calculation formula.

[0146] The five steps S5.1 to S5.5 are repeated from the start of the scheduling period to the end of the scheduling period to complete the scheduling decision according to the knowledge graph rules of the water-wind-solar hybrid system throughout the entire scheduling period.

[0147] S6, based on knowledge graph rules and graph chain reasoning method, obtains the scheduling scheme of the hydro-wind-solar hybrid power generation system calculated according to knowledge graph rules.

[0148] Example 2:

[0149] The application of this invention will be further described below with reference to specific experiments.

[0150] This invention focuses on the Yalong River Basin wind-solar-hydro pilot demonstration base, which includes four wind power stations (Wodi, Dahe, Asa, and Baiwu), one photovoltaic power station (Zhalashan), and one hydropower station (Guandi). For ease of description, Wodi, Dahe, Asa, Baiwu, Zhalashan, and Guandi are represented by the symbols W1, W2, W3, W4, S1, and H1, respectively. The installed capacities of W1, W2, W3, W4, S1, and H1 are 91.5MW, 60MW, 80MW, 99MW, 700MW, and 2400MW, respectively. The dead water level of H1 is 1328 meters, the normal storage water level is 1330 meters, and the reservoir capacity is 760 million cubic meters. The embodiment is completed using data from July 1, 2010.

[0151] This example's scheduling object knowledge graph contains 4 wind power station entities, 1 photovoltaic power station entity, and 1 hydropower station entity. The scheduling requirement is to obtain the scheduling rules for the entire day (24 hours), therefore the scheduling time period knowledge graph contains 24 entities. The inflow of H1 (Qi) and the combined output of wind power (W1, W2, W3, W4) and photovoltaic power (S1) (Nn) are used as two types of scheduling conditions, divided into 4 and 3 condition segments respectively. The outflow of H1 (Qo) is used as the variable in the scheduling decision graph.

[0152] The results analysis for the embodiments completed using the model proposed in this invention is as follows:

[0153] (a) Non-dominated front analysis of multi-objective scheduling rules:

[0154] The non-dominated frontier of the multi-objective scheduling rule optimization model in this embodiment (the multi-objective scheduling model generates multiple solutions, and the scatter plot of the objective values ​​of each solution is shown below) is as follows: Figure 5 As shown, each scatter point represents the objective value of a solution, and each solution corresponds to a scheduling rule with a different objective equilibrium point. Scheduling rule 1 represents the scheduling rule with the highest output-load matching degree, scheduling rule 50 represents the scheduling rule with the highest power generation, and rule 25 represents the scheduling rule with the most balanced two objective values. It can be seen from the figure that the two objectives have a clear inverse relationship; that is, as power generation increases, the matching degree between output and load decreases.

[0155] (b) Comparison with traditional scheduling function forms:

[0156] To verify the effectiveness of the knowledge graph rules proposed in this invention, the scheduling objectives of the knowledge graph rules are compared with those of traditional scheduling function rules, such as... Figure 6 As shown, the scheduling function rule uses inflow as the independent variable (condition) and outflow as the dependent variable (decision). Considering that there is no existing scheduling function rule for this embodiment, and also considering the fairness of the comparison, the scheduling function rule is optimized using the power generation scheduling objective of this study. The scheduling function rule for the fourth time period under the target of maximum power generation; The scheduling function rule for the fourth time period is set to achieve the highest output-load matching degree. The scheduling function rules for other time periods are similar in form, only the specific values ​​differ. The graph shows that the power generation under the knowledge graph scheduling rule is 60.89 × 10⁻⁶. 6 kWh, compared to the power generation under traditional dispatch function rules: 60.71 × 10 6 The higher kWh output and load matching ratio (0.95) of the former are also higher than that of the latter (0.943), verifying the effectiveness of the knowledge graph rules proposed in this invention.

[0157] (c) The scheduling process calculated using scheduling rules:

[0158] The scheduling process calculated using scheduling rules is as follows: Figure 7 As shown in the figure, the following analysis can be performed:

[0159] (1) From the 3rd to the 8th period, the water level of Rule 1 was the lowest in the three scheduling processes, indicating that the water head of Rule 1 was the lowest in the three scheduling processes. Therefore, the power generation of Rule 1 was also the lowest, which made the output of Rule 1 more consistent with the load. A similar phenomenon can be observed from the 12th to the 14th period. This is why Rule 1 has the highest matching degree but the lowest power generation.

[0160] (2) From the third period to the eighth period, the water level and head of Rule 50 were the highest, indicating that the power generation was also the highest. However, the excessive output exceeded the power load, which is why the power generation of Rule 50 was the highest but the matching degree was the lowest.

[0161] (3) The water level process of Rule 25 is located between the water level processes of Rule 1 and Rule 50. It has neither too much power generation nor too low a degree of matching, and is a relatively balanced operating rule.

[0162] The above results are consistent with theoretical understanding, demonstrating the correctness of the knowledge graph scheduling rules of this invention.

[0163] (d) Output stacking diagram calculated using scheduling rules:

[0164] The output stacking diagram calculated using the scheduling rules is as follows: Figure 8 As shown in the figure, the following analysis can be performed:

[0165] (1) Wind power generation is greater at night (especially during periods 19-24) than during the day (especially during periods 7-13), while photovoltaic power generation is mainly concentrated during the day (especially during periods 8-19), showing a certain degree of natural complementarity. However, relying solely on natural complementarity is far from sufficient. Hydropower stations, through their own regulation capabilities, operate in a complementary manner with wind and photovoltaic power stations to make their total output as close as possible to the power load.

[0166] (2) Rules 1, 25, and 50 are operational rules for different needs. Rule 1 focuses on the degree of matching, so its output is closest to the target. Rule 50 focuses on power generation, so its total output is higher than the load most of the time; Rule 25 is a compromise between the two, so its scheduling process is also a compromise between the two.

[0167] The above results are consistent with theoretical understanding, demonstrating the correctness of the knowledge graph scheduling rules of this invention.

[0168] (e) Knowledge graph scheduling rules:

[0169] (1) In the knowledge graph of the scheduling object, six power station entities were constructed, and relevant attributes for scheduling calculation were set. The four wind farms (Wodi, Dahe, Asa, and Baiwu) are in a "proximity" relationship. The Guandi Hydropower Station, Zhalashan Photovoltaic Power Station, and the four wind power stations are in a "complementary" relationship.

[0170] (2) A total of 24 time period entities were constructed in the scheduling time period knowledge graph. Each entity has an ID number. Time period entity 1 and time period entity 24 have “start” and “end” attributes, respectively. The time step attribute of each entity is “1 hour”, and the relationship between adjacent nodes is “immediately following”.

[0171] (3) In the scheduling situation and decision knowledge graph, condition nodes “Qi” and “Nn” are associated with time period nodes through a “link” relationship. There is a “combination” relationship between condition nodes “Qi” and “Nn”. Further expanding the condition and decision knowledge graph, the “Qi” condition is divided into 4 segments, and within each segment, the “Nn” condition is divided into 3 segments, generating a total of 12 scheduling decision combination scenarios. Each decision entity is of type “Qo”. The relationship between decision nodes and situation nodes is a “satisfaction” relationship.

[0172] Example 3:

[0173] This embodiment provides a computer program for a knowledge graph-based scheduling method for a hydro-wind-solar hybrid power generation system. This computer program is used in the aforementioned knowledge graph-based scheduling method for the hydro-wind-solar hybrid power generation system. By employing the above-described computer program, the method of this invention can be implemented through a computer program, and further, the method of this invention can be used for actual scheduling.

[0174] Implementation 4:

[0175] In this embodiment, a terminal is provided, which is equipped with at least one controller that implements any one of the knowledge graph-based hydro-wind-solar hybrid power generation system scheduling methods described above. This controller can be used to implement the entire scheduling method of this invention.

[0176] Example 5:

[0177] In this embodiment, a computer-readable storage medium is provided, including instructions that, when executed on a computer, cause the computer to perform the aforementioned knowledge graph-based hydro-wind-solar hybrid power generation system scheduling method. This ensures the portability of the method.

[0178] Example 6:

[0179] In this embodiment, a control system for a knowledge graph-based scheduling method for a hydro-wind-solar hybrid power generation system is provided. The control system is used to implement the knowledge graph-based scheduling method for the hydro-wind-solar hybrid power generation system.

[0180] Example 7:

[0181] A scheduling device for a hydro-wind-solar hybrid power generation system, incorporating a knowledge graph-based scheduling method, is disclosed. This scheduling device enhances the flexibility of its use.

[0182] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A scheduling method for a hydro-wind-solar hybrid power generation system based on a knowledge graph structure, characterized in that, The scheduling method first designs a scheduling rule structure for the hydro-wind-solar hybrid power generation system based on a "object-time period-situation-decision" knowledge graph; then, it extracts the set of variables that can be optimized from the knowledge graph rules, establishes an optimization model for the knowledge graph scheduling rules of the hydro-wind-solar hybrid power generation system, and completes the solution; finally, it obtains the scheduling scheme calculated by the hydro-wind-solar hybrid power generation system according to the knowledge graph rules by using a graph chain reasoning method based on the current scheduling situation to obtain the decision results. The specific structure of the "Object-Time Period-Situation-Decision" knowledge graph scheduling rules is as follows: (1) The scheduling object map establishes the relationship between hydropower, wind power and photovoltaic computing object entities through hydraulic topology connection, multi-site proximity relationship and water-wind-solar complementarity relationship, and assigns characteristic value, characteristic curve and scheduling constraint attribute to each computing entity; (2) The scheduling time period map mainly establishes the time period connection relationship for each scheduling time period entity, and clarifies the time period index, time period start time, time period end time, time scale, and total number of scheduling time periods; (3) The scheduling situation map establishes the situation entity for the facing period from three aspects: meteorology, water conditions and engineering conditions. The inflow, water level above the dam, wind speed or wind power output, solar radiation intensity or photovoltaic output for the facing period are used as situation attributes. The combination relationship of each scheduling situation and the closed loop relationship formed by all situations are considered. (4) The scheduling decision map completes the scheduling decision of water, wind and solar power in the time period based on the scheduling object map, scheduling time period map and scheduling situation map. It takes various decision results as entities, the water level at the end of the time period, the reservoir outflow of the time period or the hydropower output decision value of the time period as attributes, and the relationship between the situation conditions and the decision results as the relationship. Chain reasoning is completed in five steps: S5.1, Based on the scheduling object map, select the hydropower, wind power, and photovoltaic objects participating in this scheduling calculation, and determine the complementary scheduling relationship; S5.2, Select scheduling periods from the scheduling period map. t ; S5.3 First, based on the wind power and solar power dispatch map objects in step S5.1 and the time period calculation reasoning in step S5.2, the sum of wind power and solar power output for the time period is obtained. Then, based on the hydropower dispatch map objects in step S5.1 and the time period for the time period, the initial water level above the dam for the time period is obtained. S5.4, Based on the reasoning of the two-dimensional variables in the scheduling situation graph in step S5.3, the time period hydropower output decision value under this scheduling situation combination is obtained based on the scheduling decision graph reasoning. S5.5 Assign the reasoning result of step S5.4 to the hydropower scheduling object, and obtain the scheduling process of the reservoir upstream reservoir and reservoir outflow at the end of the reservoir period based on the water balance equation and hydropower output calculation formula. The five steps S5.1 to S5.5 are repeated from the start of the scheduling period to the end of the scheduling period to complete the scheduling decision based on the knowledge graph rules of the hydro-wind-solar hybrid system throughout the entire scheduling period.

2. The scheduling method for a hydro-wind-solar hybrid power generation system based on a knowledge graph structure according to claim 1, characterized in that, The scheduling method is specifically Includes the following steps: S1 collects basic attribute data, scheduling characteristic curve data, hydrological data, and power output data of each power station in the hydro-wind-solar hybrid power generation system; S2, based on the "object-time period-situation-decision" knowledge graph, design the scheduling rule structure of the hydro-wind-solar hybrid power generation system, and clarify the entities, attributes and relationships in the knowledge graph scheduling rules; S3, extract the set of variables that can be optimized from the knowledge graph scheduling rules, and use them as decision variables for the scheduling rule optimization model; S4 aims to maximize power generation and achieve the highest degree of power output matching with load. It constructs constraints for three subsystems: hydropower, wind power, and photovoltaic power generation. It establishes a knowledge graph scheduling rule optimization model for the hydro-wind-solar hybrid power generation system and solves the problem based on a multi-objective evolutionary algorithm. S5. After the knowledge graph scheduling rules of the hydro-wind-solar hybrid power generation system are optimized, a graph chain reasoning method is proposed to obtain the decision result based on the current scheduling situation. S6, based on knowledge graph rules and graph chain reasoning method, obtains the scheduling scheme of the hydro-wind-solar hybrid power generation system calculated according to knowledge graph rules.

3. The scheduling method for a hydro-wind-solar hybrid power generation system based on a knowledge graph structure according to claim 2, characterized in that, Hydro-wind-solar hybrid power generation system has , , If there are wind power stations, photovoltaic power stations, and hydropower stations, then the dispatch target map will include them. , , One wind power station entity, one photovoltaic power station entity, and one hydropower station entity; The entire scheduling calculation period has T For each time period, from 0 to ( T -1) are all entities in the scheduling time period graph, where time period 0 is the start time period entity, and time period ( T -1) represents the end segment entity, time period t and time period t The +1s indicate an "immediate" relationship; The scheduling situation graph has multiple condition types: inbound flow Qi Wind power and photovoltaic power output Nn Wind power output Nw Photovoltaic power generation output Ns Upstream water level Zu The scheduling situation graph is one or more of the above condition types, taking... M Class conditions are used as a scheduling situation graph, and each class of conditions is divided into... If there are several condition segments, then the scheduling situation has... Given a combination of scenarios, assuming the first type of condition involves inbound traffic. Qi Type, in its possible range of values China is divided into If the segment is such that the first set of condition nodes is... ,common Each scheduling situation node M There are a total of class conditions There are several scheduling situation nodes. The first group of scheduling situation nodes and time period nodes are "linked" to each other. Nodes of the same type are "closed-loop" to each other, and nodes of different types are "combined" to each other. Scheduling decision graphs also come in various types: outbound flow Qo Upstream water level Zu In this context, the scheduling decision graph can only take one type as the decision variable, and each combination of scheduling situations corresponds to a scheduling decision node. There are a total of 10 possible combinations of situations. If the outbound flow is used as the scheduling decision variable type for a scheduling decision node, then the scheduling decision node is... .

4. The scheduling method for a hydro-wind-solar hybrid power generation system based on a knowledge graph structure according to claim 2, characterized in that, In step S3, extracting the set of variables that can be optimized from the knowledge graph scheduling rules specifically involves: The scheduling object map and the scheduling time period map can be established after the scheduling requirements are determined, and they are deterministic, meaning there are no variables that need to be optimized. After the scheduling requirements are determined, only the structure of the scheduling situation graph and scheduling decision graph can be determined. However, the specific condition values ​​for each scheduling situation entity and the decision values ​​for each scheduling decision entity cannot be determined yet. These variables need to be optimized and determined. For the first... t A time period, assuming M Class conditions serve as a scheduling situation graph, and each class of conditions can be divided into... There are several conditional segments. Considering that each conditional segment has a beginning and an end, this will generate... One scheduling situation variable to be optimized. M A total of class conditions will be generated There are several scheduling situation variables to be optimized, namely: , , ..., ; M Class conditions will produce There are several scheduling situation combinations, and each scheduling situation combination corresponds to a scheduling decision variable, i.e., there are... One scheduling decision variable to be optimized: , No. t During each time period, a total of There are a total of variables to be optimized. T The scheduling period will generate One variable to be optimized.

5. The scheduling method for a hydro-wind-solar hybrid power generation system based on a knowledge graph structure according to claim 3, characterized in that, In S4, the objective, constraints, and solution method of the knowledge graph scheduling rule optimization model for hydro-wind-solar hybrid power generation systems are as follows: (1) Scheduling objective: 1) Objective 1: Maximize power generation: ；（1） ；（2） In the formula, This is the total power generation of the hydro-wind-solar hybrid power generation system; The hydro-wind-solar hybrid power generation system is the first t Effort output during each time period; and Represents the total number of scheduling periods and the duration of a single scheduling period; , and Representing the first i The first wind power station, the first j The first photovoltaic power station, the first k The first hydropower station in t Effort output during each time period; , , This represents the number of wind power stations, photovoltaic power stations, and hydropower stations in a hydro-wind-solar hybrid power generation system. 2) Objective 2: Achieving the highest degree of output-load matching. ；（3） In the formula, It refers to the degree of matching between output and load; It is the first t Electricity load for each time period; (2) Scheduling constraints: 1) Reservoir water level constraints: ；（4） In the formula, It is the first i The reservoir in the first t The upstream water level at each time period, and These are its upper and lower limits under flood control constraints; 2) Reservoir discharge constraints: ；（5） ；（6） In the formula, It is the first i The reservoir in the first t Outbound flow rate for each time period and These are its corresponding upper and lower limit constraints; outbound flow. Power generation flow and water discharge flow Composition, power generation flow Cannot exceed the maximum transmission capacity ; 3) Hydropower output constraints: ；（7） In the formula, It is the first i The reservoir in the first t The effort exerted during each period, and These are its corresponding upper and lower limit constraints; hydropower output is also constrained by the expected output curve and the NHQ curve. 4) Water balance equation: ；（8） In the formula: and They represent the first i The reservoir in the first t Storage capacity at the beginning and end of each time period; and These are the corresponding inbound and outbound flow rates. This represents water loss due to evaporation; 5) Hydraulic connections between cascade hydropower stations: ；（9） ；（10） In the formula, It is the first i The inter-regional flow rate of each reservoir's upstream river section; It is the traffic of the upstream site. Reaching the first [location / stage] through river channel evolution i The flow rate of each reservoir at that time; Is with the first i A collection of upstream stations that are hydraulically connected to the reservoir. yes Element; This represents a river channel evolution model, specifically the Muskingen model or the time-delay evolution model. 6) Wind power constraints: ；（11） ；（12） In the formula, It is the first j The first wind power station in the t Wind speed at different times, and It refers to the entry and exit wind speeds; It is the first j The first wind power station in the t The power output during any given time period cannot exceed the installed capacity of the corresponding wind power station. ; 7) Constraints on photovoltaic power generation: ； （13） In the formula, and Representing the first k The output and installed capacity of each photovoltaic power station; 8) Conveying channel capacity constraints: ； （14） In the formula, and Representing the first g The packaged output and transmission channel capacity of the multi-energy complementary system of water, wind and solar power; (3) Solution method: The optimization model for the knowledge graph scheduling rules of the hydro-wind-solar hybrid power generation system is a multi-objective optimization problem. Based on this, an arbitrary multi-objective evolutionary algorithm is used to solve it.

6. A computer program for a knowledge graph-based scheduling method for a hydro-wind-solar hybrid power generation system, characterized in that, The computer program for the knowledge graph-based hydro-wind-solar hybrid power generation system scheduling method is used to implement the knowledge graph-based hydro-wind-solar hybrid power generation system scheduling method according to any one of claims 1 to 5.

7. A terminal, characterized in that, The terminal is equipped with at least a controller that implements the knowledge graph-based hydro-wind-solar hybrid power generation system scheduling method according to any one of claims 1 to 6.

8. A computer-readable storage medium comprising instructions, which, when executed on a computer, cause the computer to perform a knowledge graph-based scheduling method for a hydro-wind-solar hybrid power generation system as described in any one of claims 1-5.

9. A control system for a knowledge graph-based scheduling method for a hydro-wind-solar hybrid power generation system, characterized in that, The control system is used to implement the knowledge graph-based hydro-wind-solar hybrid power generation system scheduling method according to any one of claims 1-5.

10. A hydro-wind-solar hybrid power generation system scheduling device that incorporates the knowledge graph-based hydro-wind-solar hybrid power generation system scheduling method of claim 9.

Images