Optimal control method for generator sets

Abstract

The present invention relates to the field of electric power technology, and in particular to a generator set optimization control method, the method comprising: step 1: constructing a generator network; step 2: converting the generator network into a double-chain representation; step 3: performing double-chain genetic optimization on the generator network represented by the double-chain, and in the optimization process, making each generator connected to other generators as much as possible; step 4: based on the optimized generator network, connecting the generators in the generator set, operating the generator set, collecting the operating data of each generator in the generator set during the operation, and feedback-adjusting the output voltage amplitude and phase angle of each generator according to the collected operating data. The present invention improves the stability, efficiency and adaptability of the generator set by dynamically adjusting the connection relationship and output voltage parameters between the generators, thereby improving the reliability and sustainability of the power supply.

Description

Technical Field

[0001] The present disclosure relates to, but is not limited to, the field of electric power technology, and in particular to a method for optimizing control of a generator set. Background Art

[0002] Generators are the core components of power systems, responsible for converting mechanical energy into electrical energy. In order to maintain the stable operation of the power system and improve the power generation efficiency, the control of the generator set is crucial. Generator set optimization control methods are an important research field, which aims to achieve the best performance of the power system by adjusting the operating parameters of the generator. In this background technology, we will explore the existing technologies of generator set control and point out the problems and challenges.

[0003] In traditional power systems, the control of generator sets is usually based on fixed operating strategies and parameter settings. These parameters include output voltage, frequency, and power factor, which usually remain unchanged under steady-state operation. A major problem with traditional methods is that they cannot adapt to the dynamic changes of the power system and cannot achieve optimal power generation performance. Under high load or fault conditions, traditional methods may not be able to quickly adjust the parameters of the generator to deal with the problem, resulting in instability and reduced efficiency of the power system.

[0004] To address the limitations of traditional methods, some advanced control methods have been proposed. These methods use advanced control algorithms and power system monitoring data to adjust the parameters of the generator in real time. For example, the model predictive control (MPC) algorithm can adjust the output of the generator by predicting the future state of the power system. Although these methods can improve performance in some cases, they are still limited by the imperfect power system model and computational complexity. Despite some progress, generator set control still faces some important issues and challenges: the operation of the power system is highly dynamic and uncertain, including load changes, faults, and instability of renewable energy. Traditional control methods often cannot adapt to such changes, resulting in system instability and inefficiency. Power systems are complex, including multiple generators, transformers, lines, and loads. Optimizing generator set control requires considering multiple variables and constraints, which increases the complexity of the problem. Summary of the invention

[0005] The present disclosure provides a generator set optimization control method, which improves the stability, efficiency and adaptability of the generator set by dynamically adjusting the connection relationship and output voltage parameters between generators, thereby improving the reliability and sustainability of power supply.

[0006] In order to solve the above problems, the technical solution of the present invention is achieved as follows:

[0007] A generator set optimization control method, the method comprising:

[0008] Step 1: Construct a generator network, consider each generator in the generator group as a node in the generator network, and the connection between generators as an edge; in the generator network, each node is connected to at most 4 other nodes and at least 1 other node;

[0009] Step 2: Convert the generator network into a double chain representation, specifically including: construct a double chain, the structure of each chain in the double chain is the same, in the first chain of the double chain, from the beginning to the end, record the mark information of each node in turn, and in the second chain, from the beginning to the end, record the mark information of other nodes connected to the corresponding node in the first chain; in the first chain, every 4 chain spaces are a unit to store the mark information of a node, and in the second chain, also every 4 chain spaces are a unit, and every 1 chain space stores the mark information of other nodes connected to the corresponding node in the first chain. If the number of nodes connected to the node in the first chain is less than 4, but N, then in the second chain, in every unit composed of 4 chain spaces, the values ​​in 4-N chain spaces are set to 0;

[0010] Step 3: Perform double-chain genetic optimization on the generator network represented by the double chain. During the optimization process, each generator is connected to other generators as much as possible to generate a new generator network represented by the double chain. Based on the new generator network represented by the double chain, the connection relationship of the nodes in the generator network is changed to obtain an optimized generator network.

[0011] Step 4: Based on the optimized generator network, the generators in the generator set are connected and the generator set is operated. During the operation, the operating data of each generator in the generator set is collected, and the output voltage amplitude and phase angle of each generator are adjusted by feedback according to the collected operating data.

[0012] Furthermore, the step 4 specifically includes:

[0013] Step 4.1: Represent the optimized generator network as a directed graph G(V,E), where V is the set of nodes and E is the set of edges;

[0014] Step 4.2: Collect the operating data of each generator in the generator group, and calculate and optimize the active power and reactive power of each node in the generator network based on the collected operating data;

[0015] Step 4.3: Based on the calculated active power and reactive power, iteratively update the output voltage amplitude and phase angle of each generator, and perform feedback regulation on each generator.

[0016] Furthermore, the active power is calculated using the following formula:

[0017]

[0018] Among them, P i is the active power of node i; V i is the voltage amplitude at node i; G ij and B ij is the admittance parameter of the conductor (i, j); θ i and θ j is the phase angle between node i and node j; M is the total number of other nodes connected to node i;

[0019] The reactive power is calculated using the following formula:

[0020]

[0021] Among them, P i is the reactive power of node i.

[0022] Furthermore, the output voltage amplitude of each generator is iteratively updated

[0023]

[0024] in, is the output voltage amplitude of node i at the kth iteration; is the output voltage amplitude of node j at the kth iteration.

[0025] Furthermore, the phase angle of each generator is iteratively updated using the following formula:

[0026]

[0027] in, is the phase angle of node i at the kth iteration; is the phase angle of node j at the kth iteration.

[0028] Furthermore, the method of performing double-chain genetic optimization on the generator network represented by the double-chain in step 3 includes:

[0029] Step 3.1: Let the double chain of the generator network be represented by Ψ init , let NUM represent the number of units in the first chain; L represents the number of chain spaces in the second chain whose values ​​in each unit are not 0; where, is the double-chain state of the sign information of the node corresponding to the k-th unit on the first chain; The double-chain state of the sign information in the v-th chain space of the unit on the second chain corresponding to the k-th unit on the first chain;

[0030] Step 3.2: Based on the crossover operator Perform a double-strand crossover operation to create new offspring;

[0031] Step 3.3: Based on mutation operator Perform double-strand mutation operation to obtain mutation results;

[0032] Step 3.4: Repeat steps 3.1 to 3.3 until the preset stop condition is reached to update the double chain and generate a generator network representing a new double chain.

[0033] Furthermore, the double chain of the generator network is represented as Ψ init :

[0034]

[0035] in, is the double-chain state of the sign information of the node corresponding to the k-th unit on the first chain; It is the double-chain state of the sign information in the v-th chain space of the unit on the second chain corresponding to the k-th unit on the first chain.

[0036] Furthermore, in step 3.2, the following formula is used to use the crossover operator: Perform a double-chain crossover operation to create new offspring

[0037]

[0038] in, for The new offspring of for The new offspring of Use the following formula to express it:

[0039]

[0040] Among them, (A) is the crossover operator The first cross factor of (B) is the crossover operator The second cross factor.

[0041] Furthermore, in step 3.3, the following formula is used based on the mutation operator Perform double-strand mutation operation:

[0042]

[0043] in, for The variation results; for The mutation result of Use the following formula to express it:

[0044]

[0045] in, is the revolving door operation; g is the variation amplitude; e is the natural base.

[0046] Furthermore, the stopping condition in step 3.4 is: the value of NUM*L exceeds the value of NUM*L in each previous iteration and exceeds the value of NUM*L in the next H iterations.

[0047] The generator set optimization control method of the present invention has the following beneficial effects: the traditional generator control method is usually based on static settings, which is difficult to cope with the dynamic changes in the power system. The present invention adopts the construction of a generator network, allowing each generator node to dynamically adjust the connection relationship with other nodes. This dynamicity allows the power system to better adapt to load changes and fault conditions, thereby improving the stability of the system. In the power system, stability is crucial because it directly affects the reliability of power supply. Through the method of genetic optimization and double chain representation, the global optimization of the generator network can be achieved. This means that the system can better match the output of the generator with the load demand, and maximize the energy utilization efficiency of the power system. Traditional control methods are usually based on fixed parameter settings, which may cause energy waste in the power system. The present invention dynamically adjusts the connection relationship between generators, so that the generators can work together in a more efficient way, thereby improving the efficiency of the entire system. Updating the output voltage amplitude and phase angle of each generator helps to improve the stability of the power system. Traditional control methods usually fix the voltage and phase angle, which is difficult to cope with the dynamic changes in the power system. The optimization control method in the present invention allows the voltage amplitude and phase angle of the generator to be dynamically adjusted according to real-time operation data to ensure that the system can remain stable under different loads and fault conditions. This real-time adjustment enhances the robustness of the power system and helps prevent power system instability or collapse. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 A schematic diagram of a method flow chart of a generator set optimization control method provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0049] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present disclosure clearer and more understandable, the present disclosure is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present disclosure and are not used to limit the present disclosure.

[0050] Example 1: Reference Figure 1 , a generator set optimization control method, the method comprising:

[0051] Step 1: Construct a generator network, treating each generator in the generator group as a node in the generator network, and the connection between generators as an edge; in the generator network, each node is connected to at most 4 other nodes and at least 1 other node; in this step, first treat each generator in the generator group as a node in the generator network, and treat the connection between generators as an edge. The key feature of this network construction is that each node is connected to at most 4 other nodes and at least 1 other node. This connection method can be regarded as a topological structure, which determines the communication and collaboration between generators.

[0052] Step 2: Convert the generator network into a double chain representation, specifically including: construct a double chain, the structure of each chain in the double chain is the same, in the first chain of the double chain, from the beginning to the end, record the mark information of each node in turn, and in the second chain, from the beginning to the end, record the mark information of other nodes connected to the corresponding node in the first chain; in the first chain, every 4 chain spaces are a unit to store the mark information of a node, and in the second chain, every 4 chain spaces are a unit, and every 1 chain space stores the mark information of other nodes connected to the corresponding node in the first chain. If the number of nodes connected to the nodes in the first chain is less than 4, but N, then in the second chain, in every unit composed of 4 chain spaces, the values ​​in 4-N chain spaces are set to 0; Compared with the traditional generator set control method, a generator network is established here, rather than simply treating the generator as an isolated unit. This network construction method can help achieve more efficient generator set collaboration, thereby improving power generation efficiency and stability. In addition, requiring each node to connect to at most 4 nodes and at least 1 node also provides a certain guarantee for the robustness and scalability of the network.

[0053] Step 3: Perform double-chain genetic optimization on the generator network represented by double chains. During the optimization process, each generator is connected to other generators as much as possible to generate a new generator network represented by double chains. Based on the generator network represented by the new double chain, the connection relationship of the nodes in the generator network is changed to obtain an optimized generator network. This step involves converting the generator network into a double chain representation. Double chain representation is a special data structure that contains two chains, each with the same structure. In this representation, the first chain is used to record the flag information of each node, and the second chain is used to record the flag information of other nodes connected to the corresponding node in the first chain. The key point of the double chain representation is to store information in units of 4 chain spaces, and in the second chain, if the number of nodes connected to the nodes in the first chain is less than 4, the corresponding chain space value is set to 0. The introduction of double chain representation is to effectively record the connection relationship between nodes in the generator network, and at the same time, by processing the situation where 4 connected nodes are not satisfied, the flexibility of the data structure is improved. In this step, the generator network represented by double chains is used for genetic optimization. Genetic optimization is a heuristic algorithm that aims to find the optimal solution. Here, the optimization goal is to make each generator connected to other generators as much as possible to generate a new double-chain representation of the generator network. Through genetic optimization, the connection relationship of the nodes in the generator network can be changed to obtain a more optimized generator network.

[0054] Step 4: Based on the optimized generator network, the generators in the generator set are connected and the generator set is operated. During the operation, the operating data of each generator in the generator set is collected, and the output voltage amplitude and phase angle of each generator are adjusted by feedback according to the collected operating data.

[0055] Specifically, the generators in the generator set are regarded as a network, in which each generator is represented as a node and the connection between the generators is represented as an edge. This network model facilitates the collaborative work and communication between the generators. Then, the double chain representation is introduced to more effectively record the connection relationship between the nodes. Next, the network is optimized by using a genetic optimization algorithm, with the goal of maximizing the performance of the entire generator set by adjusting the connection between the nodes. Finally, based on the optimized network, the operating data of each generator is collected in real time, and feedback control is performed to ensure that the output voltage amplitude and phase angle of each generator are within a stable range. By optimizing the network connection, the generators in the generator set can work together more effectively and improve the efficiency of the entire system. At the same time, real-time feedback control helps maintain the stability of the system. This method is adaptive and can adjust the network connection according to different operating conditions and load conditions to adapt to different working requirements. Through better control, unnecessary energy waste can be reduced and energy utilization can be improved.

[0056] Considering the requirement that each node is connected to at most 4 nodes, this method also increases the robustness of the system, and can maintain normal operation of the system even in the event of partial node failure. By collecting and analyzing the operating data of the generator, intelligent management and predictive maintenance can be achieved, thereby extending the life of the equipment and reducing maintenance costs.

[0057] Embodiment 2: Based on the previous embodiment, step 4 specifically includes:

[0058] Step 4.1: Represent the optimized generator network as a directed graph G(V,E), where V is the set of nodes and E is the set of edges;

[0059] Step 4.2: Collect the operating data of each generator in the generator group, and calculate and optimize the active power and reactive power of each node in the generator network based on the collected operating data;

[0060] Step 4.3: Based on the calculated active power and reactive power, iteratively update the output voltage amplitude and phase angle of each generator, and perform feedback regulation on each generator.

[0061] Specifically, step 4.1: Represent the optimized generator network as a directed graph G(V,E); The purpose of this step is to represent the optimized generator network as a directed graph, where V represents a set of nodes and E represents a set of edges. A directed graph is a graphical representation method in which the connection between each node and other nodes is directional, that is, there is a definite direction from one node to another. Representing the optimized generator network as a directed graph helps to more clearly express the connection relationship between nodes and the direction of information flow. This is crucial for subsequent calculations and control because it provides a visual representation of the system structure and helps to better understand and analyze data transmission and control in the network.

[0062] Step 4.2: Collect operating data and calculate active power and reactive power; In this step, each generator in the generator set will be monitored and its operating data will be collected, including parameters such as voltage, current, phase angle, etc. Based on these collected data, the active power and reactive power of each node (generator) are calculated. Active power and reactive power are important performance parameters in the generator set. They represent the actual power output of the generator and help determine the load and energy distribution of the system. By monitoring and calculating these parameters, the status of the system can be understood in real time and corresponding control adjustments can be made.

[0063] By updating the voltage amplitude and phase angle, the power transmission efficiency of the power system can be optimized. Specifically, adjusting the voltage amplitude and phase angle to the optimal state can help reduce transmission losses and improve the transmission efficiency of power. This is particularly important for long-distance transmission and large-scale power distribution networks, which usually face large energy losses. By updating the voltage amplitude and phase angle, load balancing and power optimization can also be achieved. In the power system, the load demands of different regions and users are constantly changing, which may lead to uneven power distribution. By adjusting the output voltage and phase angle of the generator, the match between the output of the generator and the load demand can be better matched, thereby achieving load balancing. In addition, optimizing the voltage amplitude and phase angle can also help reduce the risk of overload of transmission lines and improve the stability of the power system.

[0064] Step 4.3: Iterative adjustment based on power calculation: In this step, the output voltage amplitude and phase angle of each generator are iteratively adjusted based on the calculated active power and reactive power. This adjustment is based on the principle of feedback control, aiming to achieve the desired power output of the system. The key to this step is to achieve real-time feedback control to ensure the stability and performance of the system. By iteratively adjusting according to the actual power data, the deviation of the voltage amplitude and phase angle can be corrected in time to meet the requirements of system operation. This helps prevent problems such as voltage fluctuations, overloads, and system instability.

[0065] Embodiment 3: Based on the previous embodiment, the active power is calculated using the following formula:

[0066]

[0067] Voltage amplitude (V i ): V i Represents the voltage amplitude at node i, that is, the size of the voltage. The nodes in the generator set usually represent the connection points of the generator, load or AC line. The voltage amplitude of the node is the key parameter to represent the voltage strength, with volts as the unit. ij and B ij ): These parameters represent the electrical characteristics of the wire or cable connecting node i and node j in the generator set. ij represents the real part (conductance) of the wire, and B ij represents the imaginary part of the conductor (susceptance). They are complex numbers that describe the response of the conductor to current and voltage. i -θ j ):θ i and θ jrepresents the voltage phase angle difference between node i and node j, that is, the phase difference of the voltage waveform between the two nodes. The phase angle difference determines the relative phase of the current. Summation between nodes: The summation symbol ∑ in the formula represents the summation of other nodes j connected to node i. This means that we calculate the contribution of all nodes connected to node i.

[0068] Voltage amplitude V i Represents the voltage at node i, its square Represents the electrical energy of the voltage at node i. Admittance parameter G ij and B ij Describes the electrical characteristics of the wire, where G ij represents the component of actual power transfer, while B ij Represents the component of reactive power transmission. Phase angle difference θ i -θ j Represents the relative phase of the voltage waveform between node i and node j, which affects the direction of power transfer.

[0069] The calculation of active power takes into account the voltage amplitude, admittance parameters and phase angle difference, which determine the magnitude and direction of the current. Active power represents the actual flow direction and magnitude of electrical energy at node i.

[0070] Among them, P i is the active power of node i; V i is the voltage amplitude at node i; G ij and B ij is the admittance parameter of the conductor (i, j); θ i and θ j is the phase angle between node i and node j; M is the total number of other nodes connected to node i;

[0071] The reactive power is calculated using the following formula:

[0072]

[0073] Among them, P i is the reactive power of node i.

[0074] Voltage amplitude (V i ): V i Represents the voltage amplitude at node i, that is, the magnitude of the voltage. Voltage amplitude is an important electrical characteristic of nodes in a generator set, expressed in volts. ij and B ij ): These parameters represent the electrical characteristics of the wire or cable connecting node i and node j in the generator set. ij represents the real part (conductance) of the wire, and B ijrepresents the imaginary part of the conductor (susceptance). They are complex numbers that describe the response of the conductor to current and voltage. i -θ j ):θ i and θ j represents the voltage phase angle difference between nodes i and j, that is, the phase difference of the voltage waveform between the two nodes. The phase angle difference affects the relative phase of the current. Summation between nodes: The summation symbol ∑ in the formula represents the summation of other nodes j connected to node i. This means that we calculate the contribution of all nodes connected to node i.

[0075] Reactive power is the ineffective power transmitted in the generator set. Its calculation is based on the following principles: voltage amplitude V i Represents the voltage at node i, its square Represents the electrical energy of the voltage at node i. Admittance parameter G ij and B ij Describes the electrical characteristics of the wire, where B ij represents the component of actual power transfer, while G ij Represents the component of reactive power transmission. Phase angle difference θ i -θ j It represents the relative phase of the voltage waveform between node i and node j, which affects the direction of reactive power transfer.

[0076] Embodiment 4: Based on the previous embodiment, iteratively update the output voltage amplitude of each generator

[0077]

[0078] in, is the output voltage amplitude of node i at the kth iteration; is the output voltage amplitude of node j at the kth iteration.

[0079] Specifically, and These variables represent the output voltage amplitudes of nodes i and j at the kth iteration. They are the targets of iterative updates, and they are continuously adjusted to meet system performance requirements. α: α is a tuning parameter used to control the adjustment step size of each iteration. Its size affects the speed and stability of the adjustment. This part represents the active power (P i ) is a function of the node voltage, conductor parameters and phase angle difference in the generator set, and is used to estimate the expected value of active power. The main goal of the iterative update formula is to adjust the voltage amplitude of node i so that the estimated active power is consistent with the actual active power P iThe difference is reduced by continuously iteratively updating the voltage amplitude. To achieve.

[0080] First, by calculating the current voltage amplitude and the voltage amplitude of the adjacent nodes As well as the conductor parameters and phase angle difference, the predicted value of the active power at node i is estimated. Then, this predicted value is compared with the actual active power P i Compare and calculate the difference between the two. Finally, adjust the current voltage amplitude In order to reduce the estimation error of active power, the actual active power P i Close to the predicted value. The function of this formula for iteratively updating the voltage amplitude is as follows: By iteratively updating the voltage amplitude, the voltage stability of the generator set can be improved, ensuring that the voltage fluctuates within a stable range and preventing voltage collapse or excessive voltage. By adjusting the voltage amplitude, the distribution of active power can be controlled to ensure that each node distributes electric power according to the expected value. Iterative updates help achieve load balancing and avoid overloading some nodes while maintaining the normal operation of other nodes.

[0081] Embodiment 5: Based on the previous embodiment, the following formula is used to iteratively update the phase angle of each generator:

[0082]

[0083] in, is the phase angle of node i at the kth iteration; is the phase angle of node j at the kth iteration.

[0084] Specifically, and These variables represent the phase angles of nodes i and j at the kth iteration. They are the targets of iterative updates, and they are continuously adjusted to meet system performance requirements. α: α is a tuning parameter used to control the adjustment step size of each iteration. Its size affects the speed and stability of the adjustment. This part represents the reactive power (Q i ) is a function of the node voltage, conductor parameters and phase angle difference in the generator set, and is used to estimate the expected value of reactive power. The main goal of the iterative update formula is to adjust the phase angle of node i. To reduce the estimation error of reactive power, the actual reactive power Q i Close to the predicted value.

[0085] First, by calculating the current phase angle The phase angle between adjacent nodes And the voltage amplitude and conductor parameters, the predicted value of reactive power at node i is estimated. Then, this predicted value is compared with the actual reactive power Q i Compare and calculate the difference between the two. Finally, adjust the current phase angle To reduce the estimation error of reactive power, the actual reactive power Q i Close to the predicted value. By iteratively updating the phase angle, the voltage stability of the generator set can be improved, ensuring that the phase angle fluctuates within a stable range and preventing voltage stability problems. By adjusting the phase angle, the distribution of reactive power can be controlled to ensure that each node distributes reactive power according to the expected value. Reactive power calculation can be used in the planning of generator sets to determine reactive power requirements.

[0086] Embodiment 6: Based on the previous embodiment, the method for performing double-chain genetic optimization on the generator network represented by the double chain in step 3 includes:

[0087] Step 3.1: Let the double chain of the generator network be represented by Ψ init , let NUM represent the number of units in the first chain; L represents the number of chain spaces in the second chain whose values ​​in each unit are not 0; where, is the double-chain state of the sign information of the node corresponding to the k-th unit on the first chain; The double-chain state of the sign information in the v-th chain space of the unit on the second chain corresponding to the k-th unit on the first chain;

[0088] Step 3.2: Based on the crossover operator Perform a double-strand crossover operation to create new offspring;

[0089] Step 3.3: Based on mutation operator Perform double-strand mutation operation to obtain mutation results;

[0090] Step 3.4: Repeat steps 3.1 to 3.3 until the preset stop condition is reached to update the double chain and generate a generator network representing a new double chain.

[0091] Specifically, the double-chain crossover operation uses the crossover operator It creates a new offspring double chain based on two parent double chains. This operation simulates the crossover process in biological genetics, combining the information of the parents to produce new individuals. Through double chain crossover, new offspring can be created, which contains genetic information from both parents. The double chain mutation operation uses the mutation operator It is used to mutate a single parent double chain to generate a mutation result. The mutation operation introduces randomness and helps explore new solution spaces.

[0092] Through double-chain mutation, individuals with some changes can be obtained, and these changes may help improve the performance of the system. In steps 3.1 to 3.3, the genetic operations represented by the double chain are continuously performed to generate a new generator network represented by the double chain. These operations are repeatedly performed until the preset stop condition is reached. The stop condition is usually determined based on the performance index or the number of iterations. Finally, through the iteration of this genetic algorithm, an optimized generator network can be generated to improve the performance of the generator set.

[0093] Genetic algorithm is an optimization algorithm that simulates the biological evolution process. It searches for the optimal solution to the problem through operations such as selection, crossover and mutation. In this method, the double chain representation is used to encode the topology of the generator network. The crossover and mutation operations simulate the transmission and change of genetic information, thereby improving the performance of the generator network. Through the iteration of the genetic algorithm, a better generator network topology can be found to improve the performance and stability of the generator set. This method is suitable for complex generator network and helps to manage and optimize the topology of the system. By continuously adjusting the network structure, the performance of the system can be improved, for example, reducing transmission losses and improving voltage stability.

[0094] Example 7: Based on the previous example, the double chain of the generator network is represented as Ψ init :

[0095]

[0096] in, is the double-chain state of the sign information of the node corresponding to the k-th unit on the first chain; It is the double-chain state of the sign information in the v-th chain space of the unit on the second chain corresponding to the k-th unit on the first chain.

[0097] Specifically, this new double chain represents Ψ init The topology of the generator network is encoded in a more compact and clear way as a linear combination of double-chain states. Each double-chain state represents the connection relationship and properties between a unit (node) and other units. Using tensor product notation Will and Connect together, combine the node's mark information and connection information to construct a double-chain representation of the entire network. Weight coefficient α kvAllows to adjust the importance of each node and connection relationship. This allows different nodes and connections to be adjusted to different degrees during the optimization process of the genetic algorithm to meet specific performance goals. This new double-chain representation describes the topology of the generator network more intuitively, and each double-chain state contains the attributes of the nodes and the connection information. This makes the structure of the network easier to understand and analyze. The new double-chain representation can be directly used for genetic operations in the genetic algorithm, such as double-chain crossover and double-chain mutation. In each generation of the genetic algorithm, these double-chain states can be operated to search for a better generator network topology. Through the iteration of the genetic algorithm, the topology of the generator network can be optimized to meet the performance requirements of the generator set. For example, the performance indicators of the generator set in terms of stability, load balancing, transmission loss, etc. can be optimized.

[0098] Example 8: Based on the previous example, in step 3.2, the following formula is used to use the crossover operator Perform a double-chain crossover operation to create new offspring

[0099]

[0100] in, for The new offspring of for The new offspring of Use the following formula to express it:

[0101]

[0102] Among them, (A) is the crossover operator The first cross factor of (B) is the crossover operator The second cross factor.

[0103] Specifically, the goal of the double-chain crossover operation is to combine two parent double chains (|Ψ (A) > and

[0104] |Ψ (B) >) to create a new offspring double chain. This combination is achieved through the crossover operator The crossover operator is used to linearly combine two parent double chains, where cos(θ) and sin(θ) are weight coefficients and θ is a parameter that controls the degree of crossover. is the new child of the kth unit (node) on the first chain, by and Get by crossover. is the new child of the vth chain space on the second chain corresponding to the kth unit on the first chain, by and The crossover is obtained by crossover. The crossover angle θ controls the degree of crossover. When θ = 0, the new offspring is the same as the first parent; when θ = π / 2, the new offspring is the same as the second parent; other angles will cause the offspring to be somewhere between the two parents. The double-chain crossover operation introduces diversity because it combines two different parents to create new offspring, which helps to explore a wider solution space. This crossover operation is part of the genetic algorithm, which is used to generate new individuals to improve the topology and performance of the generator set.

[0105] One of the main functions of the crossover operator is to introduce diversity. By crossing two different parent individuals, the crossover operator can produce new individuals, which are usually between the two parents, sometimes closer to one of the parents, and sometimes completely different from the parents. This diversity helps to explore a wider solution space and avoid falling into the local optimal solution. The crossover operator linearly combines the chromosome information of the two parents. Cos(θ) and sin(θ) are weight coefficients that control the degree of crossover. Through this combination, the beneficial information of the two parents can be combined to create new individuals that may be more suitable for the problem. The parameter θ of the crossover operator can be adjusted to balance the trade-off between exploration and utilization. When θ is small, more information of the parents is retained, which helps to utilize known excellent characteristics; when θ is large, more randomness is introduced, which helps to explore new possibilities. The crossover operator simulates the genetic evolution process of organisms, in which the genetic information of the parents is passed on to the next generation. In this way, the genetic algorithm can gradually improve the individuals through generations to adapt to the needs of the problem.

[0106] Example 9: Based on the previous example, in step 3.3, the following formula is used based on the mutation operator Perform double-strand mutation operation:

[0107]

[0108] in, for The variation results; for The mutation result of Use the following formula to express it:

[0109]

[0110] in, is the revolving door operation; g is the variation amplitude; e is the natural base.

[0111] Specifically, the mutation operation is a random operation in the genetic algorithm, which aims to introduce randomness and diversity to avoid falling into the local optimal solution. The principle is to change the chromosome by applying a revolving door operation to the chromosome information of the individual, thereby generating a mutated individual. The formula includes the following elements: is a revolving door operation, which rotates the chromosomes. g is the amplitude of mutation, which controls the degree of mutation. A larger g value will introduce a larger mutation. e is a natural base number used to calculate the exponential function. By applying the mutation operator, the chromosome information is rotated, introducing a certain degree of randomness, and thus generating new individuals. This helps the genetic algorithm maintain diversity so that it can better explore the solution space. The amplitude of mutation is controlled by g, and a smaller g value will introduce a smaller mutation, while a larger g value will introduce a larger mutation. This allows a trade-off between the degree of mutation and the scope of exploration. The choice of mutation operator can be adjusted according to the needs of the specific problem to ensure that the degree of mutation adapts to the nature of the problem. Mutation is an important step in the genetic algorithm and is used together with crossover to generate the next generation of individuals. The crossover operation introduces diversity by combining the information of the parent generation, while the mutation operation introduces more randomness on this basis.

[0112] One of the main functions of mutation operation is to introduce randomness. Chromosome information is affected by the revolving door operation, which introduces a certain degree of uncertainty and helps to generate diverse individuals. These diverse individuals can help genetic algorithms avoid falling into local optimal solutions and improve the ability to search the problem solution space. The mutation amplitude g controls the degree of mutation. Smaller g values ​​introduce smaller mutations, while larger g values ​​introduce larger mutations. By adjusting the value of g, the trade-off between exploration and utilization can be balanced. Smaller g values ​​retain more original information and help to utilize known excellent characteristics; larger g values ​​introduce more randomness and help to explore new possibilities. The probability distribution of mutation operations usually follows an exponential distribution, where The g part determines the probability of mutation. Larger g values ​​lead to higher mutation probabilities, while smaller g values ​​lead to lower mutation probabilities. Mutation is an important step in genetic algorithms and is used together with crossover to generate the next generation of individuals. The crossover operation introduces diversity, and the mutation operation introduces more randomness on this basis, which helps the genetic algorithm maintain diversity and explore new possibilities when searching the solution space of the problem.

[0113] Embodiment 10: Based on the previous embodiment, the stopping condition in step 3.4 is: the value of NUM*L exceeds the value of NUM*L in each previous iteration and exceeds the value of NUM*L in the next H iterations.

[0114] Specifically, the stopping condition is a mechanism for controlling the number of iterations of the genetic algorithm. In this embodiment, the stopping condition is described as follows:

[0115] First transcendence: First, it is required that the value of NUM*L (NUM times L, where NUM represents the number of cells in the first chain and L represents the number of chain spaces with non-zero values ​​in the second chain) has increased between the current iteration and the previous iteration.

[0116] Continuous Exceedance: Then, the value of NUM*L is required to increase in H consecutive iterations. This means that the growth trend should be maintained in H consecutive iterations, not just one exceedance.

[0117] The purpose of this stopping condition is to observe whether the algorithm produces more diversity in successive iterations and whether the search results are improved with the increase in diversity. If the value of NUM*L keeps increasing in successive iterations, this may indicate that the algorithm is moving in a more promising direction, so it may make more sense to continue iterating. However, if the value of NUM*L no longer increases or starts to decrease, then stopping the iteration can avoid unnecessary calculations.

[0118] Those skilled in the art will appreciate that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices may be implemented as software, firmware, hardware, or a suitable combination thereof.

[0119] The preferred embodiments of the present disclosure are described above with reference to the accompanying drawings, but the scope of the present disclosure is not limited thereby. Any modification, equivalent substitution and improvement made by those skilled in the art without departing from the scope and essence of the present disclosure shall be within the scope of the present disclosure.

Claims

1. A generator set optimization control method, characterized in that: The method comprises: Step 1: Construct a generator network, consider each generator in the generator group as a node in the generator network, and the connection between generators as an edge; in the generator network, each node is connected to at most 4 other nodes and at least 1 other node; Step 2: Convert the generator network into a double chain representation, specifically including: construct a double chain, the structure of each chain in the double chain is the same, in the first chain of the double chain, from the beginning to the end, record the mark information of each node in turn, and in the second chain, from the beginning to the end, record the mark information of other nodes connected to the corresponding node in the first chain; in the first chain, every 4 chain spaces are a unit to store the mark information of a node, and in the second chain, also every 4 chain spaces are a unit, and every 1 chain space stores the mark information of other nodes connected to the corresponding node in the first chain. If the number of nodes connected to the node in the first chain is less than 4, but N, then in the second chain, in every unit composed of 4 chain spaces, the values ​​in 4-N chain spaces are set to 0; Step 3: Perform double-chain genetic optimization on the generator network represented by the double chain. During the optimization process, each generator is connected to other generators as much as possible to generate a new generator network represented by the double chain. Based on the new generator network represented by the double chain, the connection relationship of the nodes in the generator network is changed to obtain an optimized generator network. Step 4: Based on the optimized generator network, the generators in the generator set are connected and the generator set is operated. During the operation, the operation data of each generator in the generator set is collected, and the output voltage amplitude and phase angle of each generator are adjusted by feedback according to the collected operation data; The method for performing double-chain genetic optimization on the generator network represented by the double chain in step 3 comprises: Step 3.1: Let the double chain of the generator network be represented by Ψ init , let NUM represent the number of units in the first chain; L represents the number of chain spaces in the second chain whose values ​​in each unit are not 0; where, is the double-chain state of the sign information of the node corresponding to the k-th unit on the first chain; The double-chain state of the sign information in the v-th chain space of the unit on the second chain corresponding to the k-th unit on the first chain; Step 3.2: Based on the crossover operator Perform a double-strand crossover operation to create new offspring; Step 3.3: Based on mutation operator Perform double-strand mutation operation to obtain mutation results; Step 3.4: Repeat steps 3.1 to 3.3 until the preset stop condition is reached, thereby updating the double chain and generating a generator network represented by a new double chain; The double chain of the generator network is represented as Ψ using the following formula init : in, is the double-chain state of the sign information of the node corresponding to the k-th unit on the first chain; is the double-chain state of the sign information in the vth chain space of the unit on the second chain corresponding to the kth unit on the first chain; α kv is the weight coefficient, which is used to adjust the importance of each node and connection relationship; In step 3.2, use the following formula to use the crossover operator Perform a double-chain crossover operation to create new offspring in, for The new offspring of for The new offspring of Use the following formula to express it: Among them, (A) is the crossover operator The first cross factor of (B) is the crossover operator The second cross factor of In step 3.3, use the following formula based on the mutation operator Perform double-strand mutation operation: in, for The variation results; for The mutation result of Use the following formula to express it: in, is the revolving door operation; g is the variation amplitude; e is the natural base.

2. The generator set optimization control method according to claim 1, characterized in that: The step 4 specifically includes: Step 4.1: Represent the optimized generator network as a directed graph G(V,E), where V is the set of nodes and E is the set of edges; Step 4.2: Collect the operating data of each generator in the generator group, and calculate and optimize the active power and reactive power of each node in the generator network based on the collected operating data; Step 4.3: Based on the calculated active power and reactive power, iteratively update the output voltage amplitude and phase angle of each generator, and perform feedback regulation on each generator.

3. The generator set optimization control method according to claim 2, characterized in that: The active power is calculated using the following formula: Among them, P i is the active power of node i; V i is the voltage amplitude at node i; G ij and B ij is the admittance parameter of the conductor (i, j); θ i and θ j is the phase angle between node i and node j; M is the total number of other nodes connected to node i; The reactive power is calculated using the following formula: Among them, P i is the reactive power of node i.

4. The generator set optimization control method according to claim 3, characterized in that: Use the following formula to iteratively update the output voltage amplitude of each generator: Among them, V i (k) is the output voltage amplitude of node i at the kth iteration; V j (k) is the output voltage amplitude of node j at the kth iteration; α is a regulation parameter used to control the adjustment step size of each iteration.

5. The generator set optimization control method according to claim 4, characterized in that: Iteratively update the phase angle of each generator using the following formula: in, is the phase angle of node i at the kth iteration; V j (k) is the phase angle of node j at the kth iteration.

6. The generator set optimization control method according to claim 5, characterized in that: The stopping condition in step 3.4 is: the value of NUM*L exceeds the value of NUM*L in each previous iteration and exceeds the value of NUM*L in the next H iterations.

Images