Ship shore power and shipboard power supply coordinated control device and automatic switching method

Abstract

The present invention relates to the field of ship power system control technology, and discloses a ship shore power and shipboard power supply collaborative control device and an automatic switching method. The method includes: obtaining the operating status data of the shore power access power supply, the shipboard main power supply and the backup power supply, extracting characteristic parameters and matching to generate switching rule pairs; constructing a switching strategy model and generating standard switching instructions through hierarchical instruction encapsulation; dividing the grid area according to the power grid topology data, and using a genetic optimization algorithm to plan the optimal switching path. The device includes a controller for realizing status data collection, rule matching and instruction generation. The present invention solves the problems of low efficiency and high mismatching rate of traditional switching methods through dynamic matching and intelligent path planning; hierarchical instruction encapsulation and redundant verification improve the accuracy and reliability of instructions. The present invention significantly improves the automation level and stability of the ship power system and is suitable for complex and changeable power supply scenarios.

Description

Technical Field

[0001] The present invention relates to the technical field of ship power system control, and in particular to a coordinated control device for ship shore power and shipboard power supply and an automatic switching method. Background Art

[0002] A ship's power system is a core component for ensuring its normal operation. Its stability and reliability are directly related to the ship's safety and economic efficiency. Traditional ship power systems rely primarily on onboard generators. However, when docked in port, to reduce environmental pollution and fuel consumption, it is often necessary to switch to shore power. However, the existing technology for switching between shore power and onboard power presents numerous problems.

[0003] Traditional switching methods rely heavily on manual operations, which is not only inefficient but also prone to power outages or equipment damage due to operational errors. Although some automation technologies have been introduced, the judgment of power status and path planning during the switching process remain imprecise, making it difficult to adapt to complex and changing grid topologies and load demands. For example, existing methods lack the ability to dynamically adjust feature parameter matching and switching rule generation, leading to mismatches and switching delays.

[0004] Existing technologies underutilize grid topology data, and switching path planning is often based on static models, unable to respond in real time to load changes or power supply status fluctuations. Especially in multi-power supply coordination scenarios, the lack of comprehensive optimization of path duration, power quality, and load balancing can easily lead to voltage instability or localized overloads. Furthermore, the construction and instruction packaging processes of existing switching strategy models are relatively simplistic, without fully considering hierarchical processing and redundancy checks. This can result in logical errors or execution anomalies in the generated switching instructions.

[0005] Existing devices have limitations in control accuracy and adaptability, making them unable to meet the high-reliability power system requirements of modern ships. For example, some collaborative control devices cannot effectively eliminate mismatched rules or verify command redundancy, increasing the risk of system failure. Therefore, an automatic switching method and device that can implement dynamic matching, intelligent path planning, and hierarchical command encapsulation is urgently needed to improve the stability and automation level of ship power systems. Summary of the Invention

[0006] The object of the present invention is to provide a coordinated control device and automatic switching method for ship shore power and shipboard power supply, so as to solve the problems raised in the above background technology.

[0007] To achieve the above objectives, the present invention provides the following technical solutions: a coordinated control device for ship shore power and shipboard power supply and an automatic switching method, the method comprising:

[0008] Acquiring operating status data of multiple power sources, extracting characteristic parameters based on the operating status data, and matching the extracted characteristic parameters to obtain a switching rule pair; the multiple power sources include shore power access power, onboard main power, and onboard backup power; obtaining the switching rule pair by matching the extracted characteristic parameters includes: extracting rule descriptors at the characteristic parameters, assigning time sequence values ​​to the characteristic parameters, and finding matching points based on the rule descriptors to obtain a matching rule pair; the rule descriptors are attribute sequences that describe the characteristic parameters;

[0009] A switching strategy model is constructed based on the running status data and switching rules, and the switching strategy model is encapsulated with instructions based on the running status data to generate standard switching instructions, including:

[0010] The switching rule pairs obtained by mutual matching are used as anchor points to obtain the switching strategy model through continuous matching. Based on the sparse links obtained by matching the characteristic parameters of multiple power sources, a link segmentation method is adopted. Through expansion and screening, a complete or partial switching path model covering the target load is obtained.

[0011] Layered instruction encapsulation is adopted, and instruction encapsulation is performed through two levels. The first level encapsulates the original state space into an intermediate format that can be expressed by the protocol; the second level encapsulates the instructions in the intermediate format into the target space that needs to be executed to generate standard switching instructions.

[0012] Preferably, the method further comprises the following steps:

[0013] Obtain the grid topology data required for switching and generate a switching path based on the grid topology data; multiple power supplies collect status data based on the switching path to obtain operating status data;

[0014] The grid topology data includes the target load, link impedance, and power capacity data of the required switching area. The switching path is generated based on the grid topology data, including the following:

[0015] Identify boundaries based on target loads to build power supply areas, and plan switching paths based on the power supply areas.

[0016] And, a switching benchmark is established based on the link impedance and power capacity data, and the operation status data is collected based on the switching benchmark.

[0017] Preferably, planning the switching path according to the power supply area includes:

[0018] Constructing a two-dimensional grid topology map of the target load to be switched, and dividing the target load to be switched into a plurality of grid areas with a preset interval as a side length according to the two-dimensional grid topology map;

[0019] Based on the target load, link impedance, and power capacity data, the two-dimensional grid area is classified into three types: core power area, general power area, and redundant standby area, and the core power area is prioritized.

[0020] Randomly generate several initial switching paths to form an initial switching path set;

[0021] Constructing a fitness evaluation system, wherein the fitness evaluation system includes path time evaluation items, power quality evaluation items, and load balancing evaluation items;

[0022] The path time evaluation item is the cumulative value of the switching time between each power source in the initial switching path;

[0023] The power quality evaluation item is the sum of the voltage stability that can be maintained per unit time in each link in the initial switching path;

[0024] The load balancing evaluation items are the load distribution and the amount of power output that need to be adjusted in the initial switching path;

[0025] The initial switching path set is used as the initial population, and iterative optimization is performed according to the fitness evaluation system to obtain the optimal switching path and complete the switching path planning.

[0026] Preferably, the initial switching path set is used as the initial population, and iterative optimization is performed according to the fitness evaluation system, including:

[0027] According to the fitness evaluation system, the evaluation value of each initial switching path is calculated, and the initial switching paths are sorted from high to low according to the three evaluation items to obtain a sorted set of three initial switching paths;

[0028] According to the evaluation value, several initial switching paths are selected from the three sorted sets by probability to form three sub-populations in different directions. The probability of being selected is proportional to the evaluation value.

[0029] In each subpopulation, individual encoding, crossover operation and mutation adjustment are performed according to the genetic optimization algorithm, and unreachable path solutions are eliminated. The preset number of iterations are repeated to form three subpopulations after iteration.

[0030] Individuals are randomly selected from the three subpopulations based on their evaluation values ​​to form a basic gene pool. Three groups of crossover subpopulations are formed by crossing two of the three subpopulations. For each crossover subpopulation, the three evaluation values ​​of each individual are calculated. A multi-objective sorting is performed on the two evaluation values ​​involved in the crossover subpopulations. Based on the sorting results, individual encoding, crossover operations, and mutation adjustments are performed, and unreachable path solutions are eliminated.

[0031] After the three groups of crossover sub-populations have iterated a preset number of times, the three sub-populations are combined together, and the weighted sum of the three evaluation items is performed to obtain a comprehensive evaluation value. Individual encoding, crossover operations and mutation adjustments are performed based on the comprehensive evaluation value. After iterating a preset number of times, the individual with the largest evaluation value is obtained as the optimal switching path.

[0032] Preferably, the two evaluation values ​​involved in the crossover sub-population are subjected to multi-objective sorting, and individual encoding, crossover operation and variation adjustment are performed according to the sorting results. Before the crossover sub-population crosses, a dynamic adjustment mechanism is used to select the crossover object, wherein the selection probability of the crossover object is related to the position of the individual in the sorting, and the later the position of the individual in the sorting, the higher the probability of being selected.

[0033] Preferably, the method further comprises: assigning a value to the selection probability according to the sorting order of the individuals to be crossed, wherein the later the individual is sorted, the greater the selection probability value.

[0034] Preferably, the method further comprises the following steps:

[0035] The matched switching rule pairs are screened, and the mismatched switching rule pairs are screened. When building the switching strategy model, the model is built according to the screened switching rule pairs.

[0036] Preferably, screening the matching switching rule pairs includes: using bidirectional check and consistency verification to eliminate incorrectly matched switching rule pairs.

[0037] Preferably, the method further comprises the following steps:

[0038] The generated standard switching instructions are redundantly verified, and abnormal instructions are eliminated by comparing the execution results of the main and standby instructions.

[0039] Preferably, the present invention further includes a coordinated control device for ship shore power and shipboard power supply, which is applied to the above-mentioned method for automatic switching between ship shore power and shipboard power supply, and the device includes a controller:

[0040] The controller is used to obtain the operating status data of multiple power supplies, extract characteristic parameters according to the operating status data, and obtain switching rule pairs by matching the extracted characteristic parameters;

[0041] The controller is further configured to construct a switching strategy model according to the operating status data and the switching rule, and to perform instruction encapsulation on the switching strategy model according to the operating status data to generate a standard switching instruction.

[0042] Compared with the prior art, the present invention has the following beneficial effects:

[0043] The system uses the target resin feature acquisition module to acquire the target resin's precursor composition (such as monomer type and molecular weight distribution) and reaction condition parameters. It also indexes historical modification data, digitizing historical experience and avoiding over-reliance on manual experience, making the development of modification plans more scientific and traceable. The historical feature information clustering module clusters historical process feature information, extracting historical functional group information sets and modification parameter sets. By classifying and organizing massive amounts of historical data, it can quickly locate historical cases similar to the current target resin, providing a reference for the selection of modification parameters, shortening parameter screening time, and improving modification efficiency.

[0044] The Historical Modification Parameter Set Optimization and Screening module optimizes and sequentially screens historical parameters based on probability distribution. Using a scientific algorithm, it eliminates redundant parameters and retains highly relevant modification parameters, forming an ordered historical modification parameter sequence. This improves the reliability and practicality of the parameter sequence and provides a high-quality data foundation for subsequent trend prediction. The Modification Trend Prediction module uses sample data to construct a predictor, combining weighted calculations based on the similarity between the target process characteristics and the historical feature set. This module accurately predicts modification gain and attenuation rates, enabling technicians to grasp modification trends in advance, providing a basis for parameter adjustments and reducing trial and error costs.

[0045] The dynamic correction module analyzes the deviations between the target process characteristics and the standard information of the matching historical feature group, and dynamically adjusts the predicted gain rate and attenuation rate through the modification correction coefficient to ensure that the modification scheme can adapt to the specific conditions of the current process, improve the adaptability and accuracy of the scheme, and make the modification effect more stable. The modification scheme generation module constructs a decision maker based on the corrected gain rate and attenuation rate, and generates modification scheme prompts including parameter adjustment ratios, realizing automated prompts for modification operations, reducing manual intervention, improving the degree of automation and consistency of the modification process, and helping to improve the quality stability of photoresist products. Through in-depth mining and intelligent analysis of historical data, the system realizes the transformation from data to knowledge, providing an effective technical means for enterprises to accumulate modification experience and optimize process parameters, helping to promote the advancement of negative LDI photoresist resin modification technology and improve the production efficiency and product quality of related industries. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 This is a working principle diagram of the coordinated control device and automatic switching method for ship shore power and shipboard power supply according to the present invention;

[0047] Figure 2 Flowchart for power grid topology data processing;

[0048] Figure 3 Flowchart for switching path planning;

[0049] Figure 4 This is a design drawing of the collaborative control device. DETAILED DESCRIPTION

[0050] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0051] See also Figure 1-Figure 4 The present invention relates to a coordinated control device and automatic switching method for shore power and onboard power supply of a ship, and the specific implementation steps are as follows:

[0052] The system acquires operational status data for various power sources, including shore power, onboard main power, and onboard backup power. Feature parameters are extracted from each of these operational status data. Rule descriptors are extracted from these feature parameters. Rule descriptors are attribute sequences that describe these feature parameters. These feature parameters are assigned sequential values, and matching points are found based on the rule descriptors to obtain matching rule pairs.

[0053] A switching strategy model is constructed based on operating status data and switching rule pairs. Based on the operating status data, the switching strategy model is then encapsulated into instructions to generate standard switching instructions. Using matching switching rule pairs as anchor points, the switching strategy model is obtained through continuous matching. A link segmentation-based approach is employed to expand and filter sparse links obtained by matching characteristic parameters between multiple power sources to obtain a complete or partial switching path model covering the target load. Layered instruction encapsulation is employed, with instructions encapsulated at two levels. The first level encapsulates the original state space into an intermediate format expressible by a protocol. The second level encapsulates the instructions in the intermediate format into the target space to be executed, generating standard switching instructions.

[0054] Example 1: This example requires obtaining the grid topology data for the desired switching, generating a switching path based on the grid topology data, and collecting the operating status data of multiple power sources based on the switching path. The grid topology data includes the target load, link impedance, and power capacity data of the desired switching area.

[0055] Power supply zones are constructed based on target load boundary identification. This process requires comprehensive consideration of factors such as the target load's location, power level, and power consumption characteristics. For example, different areas of a ship, such as the engine room, cargo hold, and cockpit, have different target load distribution and power requirements. Accurately identifying the boundaries of each area is crucial to determining the appropriate power supply zone. Through detailed analysis of the target loads, it is clear which loads belong to the same power supply zone, allowing accurate power supply zones to be constructed.

[0056] After establishing the power supply area, plan the switching path based on the power supply area. When planning the switching path, fully consider the power distribution, link connectivity, and load power requirements within the power supply area. Furthermore, establish a switching benchmark based on link impedance and power capacity data. Link impedance affects power transmission efficiency and voltage stability, while power capacity determines the maximum power a power source can provide. Accurately calculating link impedance and power capacity data allows you to establish a reliable switching benchmark, providing a critical basis for subsequent operational status data collection and switching operations.

[0057] When collecting operating status data for various power sources, it's crucial to prioritize switching paths and benchmarks. For example, the shore power, onboard main power, and onboard backup power sources all reside at different points in the switching path, and their operating status data collection points and frequencies must be appropriately configured based on the benchmarks. Real-time data collection of these power sources, such as voltage, current, frequency, and power factor, allows for timely understanding of their operating status, providing accurate data support for subsequent feature parameter extraction and switching decisions.

[0058] When acquiring grid topology data, appropriate measurement methods and equipment are required to ensure data accuracy and completeness. For target load data, information such as type, power, and location must be recorded. For link impedance, the impact of factors such as line length, material, and cross-sectional area on impedance must be considered, and accurate values ​​must be obtained through measurement or calculation. For power supply capacity, parameters such as the power rating and output characteristics of the power supply must be understood.

[0059] When establishing power supply zones, complex situations may arise, such as a dispersed distribution of target loads or the presence of multiple load types. This requires careful analysis and judgment to rationally demarcate power supply zone boundaries and ensure stable and reliable power supply to loads within each zone. Furthermore, it's important to account for potential future load fluctuations and reserve sufficient expansion space to facilitate adjustments to power supply zones and switching paths as load increases or changes occur.

[0060] When planning a switching path, it's important to consider various factors, including switching reliability, switching time, and power quality. A reasonable switching path should minimize switching time and impact on the load while ensuring power supply reliability, while also ensuring that power quality meets requirements. For example, during the switching process, voltage drops and frequency fluctuations must be avoided to ensure normal load operation.

[0061] Establishing a handover baseline is a critical step, directly impacting subsequent status acquisition and handover operations. The handover baseline needs to be scientifically and rationally set based on link impedance and power supply capacity data. For example, when link impedance is high, a higher handover threshold may be required to avoid voltage instability during the handover process. When power supply capacity is low, load distribution must be optimized to prevent overload.

[0062] When collecting operational status data, it's crucial to ensure the accuracy and reliability of the data collection equipment. A comprehensive data collection and transmission system must be established to ensure timely and accurate data transmission to the control system. Furthermore, preprocessing of the collected data, such as filtering and denoising, is necessary to improve its quality and usability.

[0063] The implementation of this embodiment primarily involves acquiring grid topology data, constructing power supply areas, planning switching paths, establishing switching benchmarks, and collecting operational status data based on the switching paths and benchmarks. Each step requires careful consideration to ensure data accuracy and operational rationality, thereby laying a solid foundation for the subsequent automatic switching between shore power and onboard power. These operations enable the effective management and switching of multiple power sources, improving the reliability and stability of the ship's power supply system and meeting the ship's power requirements under varying operating conditions.

[0064] Example 2: In this example, planning switching paths based on power supply areas requires the following specific operations: A two-dimensional grid topology is constructed in the target load area to be switched. This topology displays the spatial distribution of the target loads, the locations of power access points, and the link connections in the form of plane coordinates. For example, the locations of generators, distribution cabinets, and various electrical equipment in a ship's engine room are converted into two-dimensional coordinate points based on the actual layout, and cable connection links are represented by line segments. During construction, the power parameters of each load, the physical length of each link, and the material properties must be accurately annotated to ensure that the topology is consistent with the actual grid structure.

[0065] Based on the two-dimensional grid topology, the target load to be switched is divided into several grid areas with preset intervals. The value of the preset interval is determined based on the distribution density of the target load. For example, in the densely loaded cockpit area, the interval can be set to 1 meter, so that each grid covers an area of ​​approximately 1 square meter. In the sparsely loaded cargo hold area, the interval can be adjusted to 3 meters. The topology is divided into regular grids using equidistant horizontal and vertical lines based on the load coordinates. Each grid must contain at least one load node or link segment to ensure the integrity of the regional division.

[0066] After gridding is complete, the two-dimensional grid areas are classified into three types: core power supply area, general power supply area, and redundant backup area based on target load, link impedance, and power capacity data. The core power supply area typically includes grids near the ship's main power supply. The loads within these areas are critical ship equipment, such as propulsion and navigation systems, and require extremely high power reliability, so they are designated with the highest priority. The general power supply area is where general loads, such as cabin lighting and ventilation equipment, are located, and have the next highest power priority. The redundant backup area includes grids near the shore power access point and around the ship's backup power supply. It provides backup power in the event of a main power failure, and its priority is dynamically adjusted based on the backup power supply capacity and link impedance. The classification process comprehensively evaluates the importance of the load within each grid, the transmission efficiency of the link, and the available power supply capacity. For example, if the load within a grid is radar equipment, the link impedance to the main power supply is less than 0.5Ω, and the power supply capacity remains at least 80%, it is classified as a core power supply area.

[0067] Several initial switching paths are randomly generated to form an initial switching path set. Starting from the starting power source node (such as the shore power access point), the next node is randomly selected based on grid proximity until the target load node is reached. Each path must pass through at least one grid in the core power area. For example, starting from the grid where the shore power access point is located, the path passes through the adjacent redundant backup grid, the normal power grid, and finally reaches the target load grid in the core power area, forming an initial path. The number of initial paths is determined by the scale of the power grid, generally ranging from 50 to 100, to ensure sufficient diversity in the path set.

[0068] A fitness evaluation system was constructed, which included path duration evaluation, power quality evaluation, and load balancing evaluation. The path duration evaluation item is the cumulative value of the switching time between each power source in the initial switching path. The switching time includes circuit breaker operation time and voltage synchronization adjustment time. For example, if a path contains three power source switching times, each with a switching time of 0.2 seconds, 0.3 seconds, and 0.15 seconds, the duration evaluation value is 0.65 seconds. The power quality evaluation item is the sum of the voltage stability that can be maintained per unit time for each link in the initial switching path. This is calculated by calculating the voltage deviation rate of each link (such as the proportion of time maintained within ±5% of the rated voltage) and summing them up. For example, if a path contains two links with voltage stability of 95% and 98%, respectively, the power quality evaluation value is 193%. The load balancing evaluation item is the load distribution and the number of power source outputs that need to be adjusted in the initial switching path. For example, if a path switching causes a 10% overload on the main power source, requiring adjustment of the distribution of three loads, the load balancing evaluation value is 3.

[0069] Using the initial set of switching paths as the initial population, iterative optimization is performed according to the fitness evaluation system to obtain the optimal switching path. During the iterative process, the various indicators of the evaluation system must be strictly followed, and each path must be comprehensively evaluated to select the more optimal path for subsequent operations. Through continuous optimization, the path is optimized in terms of time consumption, power quality, and load balancing. Finally, the optimal switching path is determined and the switching path planning is completed. The entire planning process must be closely integrated with the actual parameters and load requirements of the power grid to ensure that the planned switching path is practical and effective, can play a good role in the ship's power supply switching, and ensure the stable operation of the ship's power supply system.

[0070] When dividing the grid area, it is necessary to repeatedly check the accuracy of the topology map to avoid errors in area division due to coordinate deviations. When classifying the grid, it is necessary to fully consider the load changes under different working conditions of the ship, such as the difference in core load under navigation status and berthing status, and adjust the classification results in time. When generating the initial path, it is necessary to avoid path duplication or invalid cycles to ensure that each path can be actually applied to power switching. When constructing the fitness evaluation system, the calculation method of each evaluation item must be consistent with the actual grid parameter measurement method to ensure the authenticity of the evaluation results. During the iterative optimization process, the number of iterations and screening criteria must be reasonably set to avoid insufficient optimization due to too few iterations, and to prevent too many iterations from affecting planning efficiency.

[0071] Example 3: In this embodiment, the process of iterative optimization using the initial switching path set as the initial population needs to be specifically implemented according to the following steps: According to the path time evaluation item, power quality evaluation item and load balancing evaluation item in the fitness evaluation system, the evaluation values ​​of each initial switching path in these three dimensions are calculated respectively. For example, for a certain initial path, the cumulative value of the switching time between its power sources is 0.8 seconds, which corresponds to the evaluation value of the path time evaluation item; the sum of the voltage stability that can be maintained within the unit time of each link is 185%, which corresponds to the evaluation value of the power quality evaluation item; the number of load distribution and power output that need to be adjusted is 2, which corresponds to the evaluation value of the load balancing evaluation item. The calculation must strictly follow the definition of each evaluation item, combined with the grid topology data and power supply operation status data, to ensure the accuracy of the evaluation value.

[0072] The initial switching paths are sorted from high to low based on the three evaluation criteria, resulting in three sorted sets of initial switching paths. Taking the path duration evaluation criterion as an example, all initial paths are sorted from minimum (i.e., shortest switching time) to maximum (longest switching time) to form the first sorted set. Similarly, the power quality evaluation criteria are sorted from maximum (highest total voltage stability) to minimum (lowest total voltage stability) to form the second sorted set. The load balancing evaluation criteria are sorted from minimum (fewest loads and power outputs requiring adjustment) to maximum (most required adjustments) to form the third sorted set. During the sorting process, it is crucial to ensure that the order of paths in each sorted set accurately reflects their performance in the corresponding evaluation criteria.

[0073] Based on the evaluation value, several initial switching paths are probabilistically selected from each of the three sorted sets, forming three subpopulations in different directions. The probability of selection is proportional to the evaluation value; that is, the higher the evaluation value (the higher the position) in the sorted set, the higher the probability of selection. For example, in the sorted set for path duration, the probability of selection for the top 10% of paths is set to 0.8, the probability for the top 20% is set to 0.6, and so on. Through linear or nonlinear probability distribution, paths with higher evaluation values ​​have a greater chance of entering the subpopulation. The number of paths selected for each subpopulation is determined by the size of the initial population, typically around 30% of the initial number of paths, to ensure a certain degree of diversity and representativeness in the subpopulation.

[0074] Within each subpopulation, individual encoding, crossover, and mutation adjustments are performed according to a genetic optimization algorithm, and unreachable path solutions are eliminated. During individual encoding, the node sequence of each path is converted into a genetic code, for example, a numerical sequence representing the order of grid areas traversed by the path. Crossover operations use a single-point or multi-point crossover method to randomly select code segments from two parent paths and exchange them to generate new child paths. Mutation adjustments randomly change a specific genetic bit in the code, such as replacing a node in the path, to increase population diversity. During this process, the newly generated path is checked for reachability, that is, whether there is a valid link from the starting power node to the target load node. If unreachable, the path solution is eliminated. Each subpopulation is iterated a preset number of times, typically 50 to 100 times, allowing the subpopulation to gradually evolve towards a more optimal path through multiple iterations.

[0075] After completing the initial iteration of the subpopulations, individuals are randomly selected from the three subpopulations based on their evaluation scores to form the base gene pool. During selection, individuals with higher evaluation scores within each subpopulation are more likely to be selected, but a certain degree of randomness must be maintained to avoid a homogenized gene pool. For example, individuals with the top 50% evaluation scores from each subpopulation are randomly sampled with a certain probability to form the base gene pool containing the dominant genes from each subpopulation.

[0076] Based on the basic gene pool, three subpopulations are crossed in pairs to form three groups of crossover subpopulations. For example, subpopulation A crosses with subpopulation B to form crossover subpopulation AB, subpopulation B crosses with subpopulation C to form crossover subpopulation BC, and subpopulation C crosses with subpopulation A to form crossover subpopulation CA. For each crossover subpopulation, three evaluation values ​​are calculated for each individual. Then, a multi-objective ranking is performed on the two evaluation values ​​involved in the crossover subpopulation. For example, crossover subpopulation AB involves path duration and power quality evaluation. These two evaluation values ​​require a comprehensive ranking. This ranking method can adopt a multi-objective ranking strategy based on Pareto optimality, classifying individuals into different levels based on their performance on the two evaluation values.

[0077] Based on the sorting results, the crossover subpopulations undergo individual encoding, crossover, and mutation adjustments. This operation is similar to genetic operations within the subpopulations, but adjustments are made based on the multi-objective sorting results to balance the optimization objectives of the two evaluation criteria. After completion, unreachable path solutions are eliminated to ensure that all paths in the crossover subpopulations are valid. Each of the three crossover subpopulations is iterated a preset number of times to further optimize the paths' overall performance across the two evaluation criteria.

[0078] After the three crossover subpopulations have been iterated a preset number of times, the three subpopulations are combined and a weighted sum of the three evaluation items is taken to obtain a comprehensive evaluation value. Weighting coefficients are set based on the actual needs of the ship's power supply system, such as a weight of 0.4 for path duration, 0.3 for power quality, and 0.3 for load balancing. A weighted summation is used to calculate the comprehensive evaluation value for each path. Based on the comprehensive evaluation value, individual encoding, crossover operations, and mutation adjustments are performed. The path is then optimized again using a genetic optimization algorithm and iterated a preset number of times. The individual with the highest evaluation value is ultimately determined as the optimal switching path.

[0079] Throughout the iterative optimization process, the parameter settings for each step must be strictly controlled, such as the proportion of probabilistic selection, the probability of genetic operations, and the distribution of weighted coefficients. These parameters must be reasonably adjusted based on the scale of the power grid, the characteristics of the load, and the configuration of the power supply. At the same time, it is necessary to ensure that each operation is based on accurate evaluation data and reliable path information to avoid optimization results deviating from actual requirements due to data errors or operational errors. Through this multi-level, multi-objective iterative optimization process, the switching path with the best overall performance in terms of time consumption, power quality, and load balancing can be screened from the initial switching path set. This provides a scientific and reasonable path planning solution for the automatic switching between shore power and onboard power supply, ensuring that the switching process is efficient, stable, and meets power supply quality requirements.

[0080] Example 4: In the process of performing multi-objective sorting and genetic manipulation on the crossover subpopulations in Example 3, a dynamic adjustment mechanism is required to select crossover targets before the crossover subpopulations cross. The core of this mechanism is to make the selection probability of the crossover target correlated with the position of the individual in the sorting. The later the individual is in the sorting, the higher the probability of being selected. At the same time, the selection probability is assigned according to the sorting order of the individuals to be crossed. The following detailed implementation method is described with reference to specific examples.

[0081] Suppose that in a crossover subpopulation (e.g., one involving both path time and load balancing), a multi-objective sorting process yields a sorted set of 10 individuals, labeled individuals 1 to 10 in descending order of their comprehensive evaluation performance. Individual 1 has a switching time of 0.5 seconds in the path time evaluation and requires one load adjustment in the load balancing evaluation. Individual 10 has a switching time of 1.2 seconds and requires four load adjustments. In this case, individuals with lower rankings (e.g., individual 10) should have a higher probability of being selected during crossover than individuals with higher rankings (e.g., individual 1).

[0082] When assigning values, either linear or nonlinear functions can be used to determine the selection probability. For example, using a linear assignment method, the selection probability of individual 1, ranked highest, is set to 0.1, that of individual 2, ranked second, to 0.15, and so on, increasing with the position in the ranking. The selection probability of individual 10 is set to 0.95. This assignment method gives individuals ranked lower in the ranking a higher chance of participating in the crossover operation. Taking the crossover operation of this subpopulation as an example, when selecting two parents from 10 individuals for the crossover, the probability of individual 10 being selected is 0.95, and the probability of individual 1 being selected is 0.1. The system uses a random number generator to generate a random number between 0 and 1. If the random number is less than 0.95, individual 10 is more likely to be selected; otherwise, other individuals may be selected.

[0083] During the crossover operation, assume that individuals 10 (ranked 10) and 7 (ranked 7, with a probability of 0.7) are selected as parents. Individual 10's genetic code is [shore power access point → redundant backup grid A → normal power grid B → target load in the core power area], and individual 7's genetic code is [shipboard main power → normal power grid C → redundant backup grid D → target load in the core power area]. Using a single-point crossover method, crossover is performed randomly at the third genetic position (between normal power grid B and redundant backup grid D). The resulting offspring 1 has the genetic code [shore power access point → redundant backup grid A → redundant backup grid D → target load in the core power area], and offspring 2 has the genetic code [shipboard main power → normal power grid C → normal power grid B → target load in the core power area].

[0084] After generating a child, the path's reachability must be checked. For example, in the path of child 1, is there a valid link from redundant backup area grid A to redundant backup area grid D? If so, the path is retained; if not, the child path is discarded. Assuming there is a cable connection between grids A and D, child 1 is a valid path.

[0085] In the mutation adjustment phase, a mutation operation is performed on the offspring 1 generated by the crossover, randomly changing one of the gene bits. For example, the redundant backup area grid D is mutated to the ordinary power supply area grid E, resulting in a new path [shore power access point → redundant backup area grid A → ordinary power supply area grid E → core power area target load]. The path is then checked for reachability. If a link exists between grids A and E, the mutated path is retained.

[0086] Through this dynamic adjustment mechanism, during the iterative process of the crossover subpopulation, individuals with lower rankings, due to their higher selection probability, participate more frequently in crossover and mutation, thereby introducing potential advantages in their genes (such as a unique path structure on a certain evaluation item) into the population. For example, although individual 10 is ranked low in the comprehensive ranking, it may have a path segment on the load balancing evaluation item that reduces the number of load adjustments. Through multiple crossover operations, this advantageous path segment may be combined with the excellent genes of other individuals (such as path segments with shorter time consumption), generating a new individual that performs better in multi-objective evaluations.

[0087] In practical applications, the selection probability assignment needs to be adjusted based on the size of the crossover subpopulation and the differences in the evaluation criteria. For example, if a crossover subpopulation contains 20 individuals, the selection probability of the top 5 individuals can be set to 0.1-0.3, the middle 10 to 0.4-0.7, and the bottom 5 to 0.8-1.0, giving individuals with lower rankings a significantly higher chance of crossover. Furthermore, within multiple crossover subpopulations (e.g., AB, BC, and CA), each group must independently assign a selection probability based on the ranking results of the corresponding evaluation criteria, ensuring that each crossover operation is optimized for a specific evaluation objective.

[0088] Furthermore, the dynamic adjustment mechanism must work in conjunction with other operations within the genetic algorithm. For example, during mutation adjustment after the crossover operation, the mutation probability of individuals ranked lower can be appropriately increased, further increasing their genetic diversity. For example, if the mutation probability of individual 10 is set to 0.3 and the mutation probability of individual 1 is set to 0.1, higher mutation probabilities will allow individuals ranked lower to generate more new path combinations, thereby expanding the search space and preventing the algorithm from falling into local optima.

[0089] During the iteration process, as the crossover subpopulation continues to evolve, the ranking of individuals changes dynamically, and the selection probability must be re-assigned based on each ranking result. For example, an individual ranked low in the first iteration may become high in the subsequent iteration due to genetic recombination, and its selection probability will decrease accordingly, and vice versa. This dynamic adjustment ensures that the selection probability is always relevant to the individual's current ranking position, maintaining the algorithm's search vitality.

[0090] Through the above-mentioned dynamic adjustment mechanism, in the optimization process of the crossover sub-population, the optimization requirements of different evaluation items can be effectively balanced, avoiding the reduction of the search space due to excessive focus on individuals with higher rankings. At the same time, by giving individuals with lower rankings a higher crossover probability, the potential high-quality path combinations can be fully explored, and the efficiency and quality of iterative optimization can be improved, which ultimately helps to generate a better switching path in the multi-objective evaluation and ensure that the ship power switching process achieves better comprehensive performance in terms of time consumption, power quality and load balancing.

[0091] Embodiment 5: This embodiment includes operations of screening the matching switching rule pairs and performing redundancy verification on the generated standard switching instructions. The following describes the implementation in detail with reference to specific examples.

[0092] To screen matching switching rule pairs, bidirectional verification and consistency validation are used to eliminate incorrectly matched rule pairs. Assume that in a ship power supply scenario, after matching the characteristic parameters of the shore power source and the shipboard main power source, a set of switching rule pairs is generated. The rule descriptor contains a sequence of attributes such as voltage amplitude, frequency, and phase difference. During bidirectional verification, verification is first performed from the shore power source to the shipboard main power source, checking whether the shore power voltage amplitude is within the allowable range of the shipboard main power source, whether the frequency deviation is less than 0.5Hz, and whether the phase difference is within ±10 degrees. Then, verification is performed in the reverse direction from the shipboard main power source to the shore power source, confirming that the characteristic parameters of the shipboard main power source meet the shore power access requirements. If the shore power voltage amplitude is 400V in the forward verification, and the allowable range of the shipboard main power source is 380V-420V, the conditions are met; and if the shipboard main power source frequency is 50Hz in the reverse verification, and the allowable range of the shore power source is 49.5Hz-50.5Hz, the conditions are also met, then the rule pair passes bidirectional verification.

[0093] Consistency verification requires comparing whether the descriptions of different power supplies on the same characteristic parameters are consistent. For example, the characteristic parameters of the shipboard main power supply and the shipboard backup power supply both include the link impedance attribute. The link impedance description for the main power supply is "the cable impedance from the main distribution board to the target load is 0.2Ω", and the description for the backup power supply is "the cable impedance from the backup power cabinet to the target load is 0.25Ω". There are differences in the description of the link impedance between the two, but these are normal attribute differences of different power supplies and are not considered inconsistent. If the voltage frequency of the shore power access power supply in a certain rule pair is described as "50Hz±0.2Hz", and the frequency of the shipboard main power supply is described as "50Hz±0.5Hz", there is a contradiction in the description of the allowable frequency deviation between the two, which may cause frequency synchronization failure during switching. At this time, the rule pair will be eliminated by consistency verification.

[0094] For example, let's consider a mismatched rule pair. For example, suppose the phase difference rule descriptor for the shore power source's characteristic parameters is "allowed phase difference of ±15 degrees," while the phase difference rule descriptor for the onboard main power source is "allowed phase difference of ±10 degrees." During matching, the allowed phase difference range for both was mistakenly recorded as ±15 degrees, resulting in a phase difference matching point of 15 degrees in the generated switching rule pair. During bidirectional verification, when switching from shore power to the onboard main power source, the actual allowed phase difference for the onboard main power source is ±10 degrees. 15 degrees exceeds this range, resulting in verification failure. This rule pair is identified as a mismatch and rejected.

[0095] After the correct switching rule pairs are screened, they are used to build a switching strategy model. For example, a correct rule pair includes switching between shore power and the ship's main power source when the voltage amplitude deviation is ≤5%, the frequency deviation is ≤0.3Hz, and the phase difference is ≤8 degrees. Based on these rule pairs and using them as anchor points, a switching path model from shore power access to the ship's main power source is found through continuous matching to ensure the accuracy of the switching strategy.

[0096] During redundancy verification of the generated standard switching instructions, abnormal instructions are eliminated by comparing the execution results of the primary and backup instructions. Assume that the generated primary instruction is "Disconnect the shore power circuit breaker, delay 0.5 seconds, then close the shipboard main power circuit breaker," and the backup instruction is "Disconnect the shore power circuit breaker, delay 0.6 seconds, then close the shipboard main power circuit breaker." When these two instructions are executed in a simulated switching scenario, when the primary instruction is executed, the shore power circuit breaker is disconnected for 0.1 seconds, and the shipboard main power circuit breaker is closed after a 0.5-second delay. The entire switching process takes 0.6 seconds, and the load voltage stabilizes at 390V and the frequency at 50Hz. When the backup instruction is executed, the shore power circuit breaker is disconnected for 0.1 seconds, and the shipboard main power circuit breaker is closed after a 0.6-second delay. The switching process takes 0.7 seconds, and the load voltage is 395V and the frequency is 50Hz. Both execution results are normal, with voltage and frequency within normal ranges. Therefore, both the primary and backup instructions are valid and retained in the instruction set.

[0097] If the load voltage drops to 350V for more than 0.2 seconds after the primary command is executed, while the voltage stabilizes at 390V after the backup command is executed, the primary command execution result is abnormal and will be eliminated through redundancy verification, while the backup command is retained. For another example, if the primary command fails to close the ship's main power circuit breaker, but the backup command executes successfully, the primary command can also be judged as abnormal and eliminated.

[0098] In practice, the primary and backup instructions for redundancy verification often have subtle differences in parameter settings, such as delay time and circuit breaker operation sequence, to account for different abnormal situations. For example, while the primary instruction performs switchover according to the normal process, the backup instruction adds 0.1 second to the primary instruction delay time to prevent switchover failures caused by circuit breaker delays. During verification, the primary and backup instructions must be executed under the same grid topology and power supply operating conditions to ensure the reliability of the comparison results.

[0099] In addition to bidirectional validation and consistency verification, switching rule pairs must also be screened against grid topology data. For example, if a switching rule requires switching from shore power to onboard backup power to pass through a certain link, but that link is shown as faulty in the grid topology, the rule pair will be rejected due to link unreachability, even if the characteristic parameters match.

[0100] When building a switching strategy model, the selected switching rule pairs serve as anchor points to ensure that the switching paths in the model are generated based on the correct rules. For example, if the rule pair determines the switching conditions for power source A to power source B, the model will use these conditions as the basis to find a feasible path from power source A to power source B, and then expand it to a complete switching path that covers the target load.

[0101] Redundancy verification also needs to consider the timing of command execution. For example, the timing of the main command "open shore power circuit breaker" and "close shipboard main power circuit breaker" must be correct. If the main command closes the main power circuit breaker before opening the shore power circuit breaker, it will cause a power short circuit. However, if the backup command timing is correct, the main command will be judged as abnormal and rejected.

[0102] The above rigorous screening of switching rule pairs and redundant verification of standard switching instructions ensures the reliability and stability of the automatic switching system between shore power and onboard power. Screening out mismatched rule pairs prevents power switching failures or equipment damage caused by incorrect switching conditions. Redundant verification eliminates abnormal instructions, ensuring reliable command execution under various operating conditions, thereby ensuring the safe and stable operation of the ship's power supply system and meeting the ship's power switching needs in different scenarios.

[0103] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0104] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for automatically switching between shore power and onboard power, characterized in that: Includes the following: Acquiring operating status data of multiple power sources, extracting characteristic parameters based on the operating status data, and matching the extracted characteristic parameters to obtain a switching rule pair; the multiple power sources include shore power access power, onboard main power, and onboard backup power; obtaining the switching rule pair by matching the extracted characteristic parameters includes: extracting rule descriptors at the characteristic parameters, assigning time sequence values ​​to the characteristic parameters, and finding matching points based on the rule descriptors to obtain a matching rule pair; the rule descriptors are attribute sequences that describe the characteristic parameters; A switching strategy model is constructed based on the running status data and switching rules, and the switching strategy model is encapsulated with instructions based on the running status data to generate standard switching instructions, including: The switching rule pairs obtained by mutual matching are used as anchor points to obtain the switching strategy model through continuous matching. Based on the sparse links obtained by matching the characteristic parameters of multiple power sources, a link segmentation method is adopted. Through expansion and screening, a complete or partial switching path model covering the target load is obtained. Adopting layered instruction encapsulation, it performs instruction encapsulation in two levels. The first level encapsulates the original state space into an intermediate format that can be expressed by the protocol. The second level encapsulates the instructions in the intermediate format into the target space to be executed, generating standard switching instructions. Also included: Obtain the grid topology data required for switching and generate a switching path based on the grid topology data; multiple power supplies collect status data based on the switching path to obtain operating status data; The grid topology data includes the target load, link impedance, and power capacity data of the required switching area. The switching path is generated based on the grid topology data, including the following: Identify boundaries based on target loads to build power supply areas, and plan switching paths based on the power supply areas. and, establishing a switching benchmark based on link impedance and power capacity data, and collecting operating status data based on the switching benchmark; Planning switching paths based on power supply areas includes: Constructing a two-dimensional grid topology map of the target load to be switched, and dividing the target load to be switched into a plurality of grid areas with a preset interval as a side length according to the two-dimensional grid topology map; Based on the target load, link impedance, and power capacity data, the two-dimensional grid area is classified into three types: core power area, general power area, and redundant standby area, and the core power area is prioritized. Randomly generate several initial switching paths to form an initial switching path set; Constructing a fitness evaluation system, wherein the fitness evaluation system includes path time evaluation items, power quality evaluation items, and load balancing evaluation items; The path time evaluation item is the cumulative value of the switching time between each power source in the initial switching path; The power quality evaluation item is the sum of the voltage stability that can be maintained per unit time in each link in the initial switching path; The load balancing evaluation items are the load distribution and the amount of power output that need to be adjusted in the initial switching path; The initial switching path set is used as the initial population, and iterative optimization is performed according to the fitness evaluation system to obtain the optimal switching path and complete the switching path planning.

2. The method for automatically switching between shore power and onboard power supply according to claim 1, characterized in that: Using the initial switching path set as the initial population, iterative optimization is performed according to the fitness evaluation system, including: According to the fitness evaluation system, the evaluation value of each initial switching path is calculated, and the initial switching paths are sorted from high to low according to the three evaluation items to obtain a sorted set of three initial switching paths; According to the evaluation value, several initial switching paths are selected from the three sorted sets by probability to form three sub-populations in different directions. The probability of being selected is proportional to the evaluation value. In each subpopulation, individual encoding, crossover operation and mutation adjustment are performed according to the genetic optimization algorithm, and unreachable path solutions are eliminated. The preset number of iterations are repeated to form three subpopulations after iteration. Individuals are randomly selected from the three subpopulations based on their evaluation values ​​to form a basic gene pool. Three groups of crossover subpopulations are formed by crossing two of the three subpopulations. For each crossover subpopulation, the three evaluation values ​​of each individual are calculated. A multi-objective sorting is performed on the two evaluation values ​​involved in the crossover subpopulations. Based on the sorting results, individual encoding, crossover operations, and mutation adjustments are performed, and unreachable path solutions are eliminated. After the three groups of crossover sub-populations have iterated a preset number of times, the three sub-populations are combined together, and the weighted sum of the three evaluation items is performed to obtain a comprehensive evaluation value. Individual encoding, crossover operations and mutation adjustments are performed based on the comprehensive evaluation value. After iterating a preset number of times, the individual with the largest evaluation value is obtained as the optimal switching path.

3. The method for automatically switching between shore power and onboard power supply according to claim 2, characterized in that: The two evaluation values ​​involved in the crossover sub-population are sorted by multiple objectives, and individual encoding, crossover operation and variation adjustment are performed according to the sorting results. Before the crossover sub-population crosses, a dynamic adjustment mechanism is used to select the crossover object, wherein the selection probability of the crossover object is related to the position of the individual in the sorting. The later the position of the individual in the sorting, the higher the probability of being selected.

4. The method for automatically switching between shore power and onboard power supply according to claim 3, characterized in that: Also includes: The selection probability is assigned according to the sorting order of the individuals to be crossed. The later the individual is sorted, the greater the selection probability value.

5. The method for automatically switching between shore power and onboard power supply of a ship according to claim 1, characterized in that: Also included: The matched switching rule pairs are screened, and the mismatched switching rule pairs are screened. When building the switching strategy model, the model is built according to the screened switching rule pairs.

6. The method for automatically switching between shore power and onboard power supply according to claim 5, characterized in that: The matching switching rule pairs are screened by using bidirectional checksum and consistency verification to eliminate incorrectly matched switching rule pairs.

7. The method for automatically switching between shore power and onboard power according to any one of claims 1 to 6, characterized in that: Also included: The generated standard switching instructions are redundantly verified, and abnormal instructions are eliminated by comparing the execution results of the main and standby instructions.

8. A coordinated control device for ship shore power and onboard power supply, applied to the method for automatic switching between ship shore power and onboard power supply according to any one of claims 1 to 7, comprising a controller, characterized in that: The controller is used to obtain the operating status data of multiple power supplies, extract characteristic parameters according to the operating status data, and obtain switching rule pairs by matching the extracted characteristic parameters; The controller is further configured to construct a switching strategy model according to the operating status data and the switching rule, and to perform instruction encapsulation on the switching strategy model according to the operating status data to generate a standard switching instruction.

Images