Large-scale oil-filled electrical equipment sensor optimization arrangement method and system

Abstract

The invention discloses a large-scale oil-filled electrical equipment sensor optimization arrangement method and system, and the method comprises the steps: building a three-dimensional model of large-scale oil-filled electrical equipment, and simulating a plurality of typical fault signal sources; for each fault signal source, an improved firefly algorithm is adopted to carry out iterative optimization to obtain an optimal sensor arrangement scheme under the fault, the improved firefly algorithm adopts a weighted multi-factor objective function as fitness, and a penalty function based on distance and a Pareto archiving strategy are introduced to promote multi-polarity optimization; and finally, comprehensively analyzing the data sets of all the fault types, outputting a final sensor arrangement scheme considering all the fault types, and identifying feature identification sensors of different faults. According to the method, the problems of poor pertinence and inaccurate measurement of a traditional sensor arrangement mode are solved, the multi-polarity characteristic of the improved firefly algorithm is utilized, the sensor layout is optimized for various fault types, and the fault monitoring precision and the early warning reliability are remarkably improved.

Description

Technical Field

[0001] This invention belongs to the field of electrical equipment monitoring technology, specifically relating to a method and system for optimizing the arrangement of sensors in large oil-filled electrical equipment. Background Technology

[0002] Large oil-filled electrical equipment, such as large power transformers and instrument transformers, are the core hubs of the power system. Their internal structures are complex, and they are susceptible to various types of potential faults during operation, such as localized overheating, partial discharge, and winding deformation. To provide timely warnings and diagnose these faults, it is typically necessary to place various types of sensors on the surface or inside the equipment, such as temperature sensors and ultrasonic sensors (for monitoring partial discharge).

[0003] Traditional sensor deployment methods often rely on engineering experience or simple simulations, lacking systematic optimization. Existing wireless sensor network (WSN) optimization algorithms, such as the Firefly Algorithm (FA), are mostly focused on achieving maximum network coverage or minimum power consumption, without being optimized for the complex multi-failure modes of large electrical equipment.

[0004] Specifically, different faults (such as localized overheating points in different locations) have different signal sources (such as heat sources or sound sources) located within the equipment. Traditional "wide-coverage" placement schemes may result in sensors being too far from the signal sources of some critical faults, leading to weak signal strength, low signal-to-noise ratio, inaccurate measurement data, and difficulty in effectively identifying fault characteristics and providing early warnings. Therefore, it is necessary to provide a new optimized placement method to address the problems of insufficient measurement accuracy and low fault identification in traditional sensor placement schemes. Summary of the Invention

[0005] To address the aforementioned shortcomings in existing technologies, this invention provides a method and system for optimizing the arrangement of sensors in large oil-filled electrical equipment, which solves the problems of low measurement accuracy and poor fault identification capability in traditional sensor arrangement methods.

[0006] To achieve the above-mentioned objectives, the technical solution adopted by this invention is as follows: a method for optimizing the arrangement of sensors in large oil-filled electrical equipment, comprising the following steps:

[0007] S100. Establish a three-dimensional model of large oil-filled electrical equipment and set the fault signal emission source under various fault modes;

[0008] S200: Simulate the fault signals in the corresponding fault modes sequentially through the fault signal transmitter, and optimize the sensor arrangement scheme when collecting fault signals by executing the improved firefly algorithm to obtain the optimal sensor arrangement scheme for each fault mode.

[0009] In the improved firefly algorithm: the fitness of the firefly is calculated and corrected by a weighted multi-factor objective function and a penalty function. When updating the spatial position of the firefly based on the fitness, the updated spatial position of the firefly is determined by combining an elite strategy and engineering constraints.

[0010] S300: The optimal sensor layout scheme under various fault modes is comprehensively analyzed and processed to obtain the globally optimal sensor layout scheme with marked common optimal sensors and feature recognition sensors.

[0011] Further, in step S200, for any fault mode, the improved firefly algorithm is executed to optimize the sensor arrangement scheme when collecting fault signals, including:

[0012] S201. Based on the fault signal emission source corresponding to the fault mode, simulate the fault signal when the fault mode is generated.

[0013] S202. Within the feasible area around the 3D model, a firefly population is randomly generated as the initial sensor layout scheme, and each firefly corresponds to a sensor node and has a unique spatial location.

[0014] S203. Collect the fault signal generated by the fault signal transmitter received by each firefly at its current spatial location;

[0015] S204. Based on the fault signals received by each firefly, calculate and correct its fitness using a weighted multi-factor objective function.

[0016] S205. Based on the fitness of each firefly, determine the candidate spatial locations for firefly updates, and combine elite strategies and engineering constraints to determine the updated spatial locations of fireflies.

[0017] S206. Repeat steps S203 to S205 until the termination condition is met, and output the optimal sensor layout scheme corresponding to the current fault mode.

[0018] Further, in step S204, the weighted multi-factor objective function is expressed as:

[0019] ;

[0020] In the formula, Fireflies Adaptability, Fireflies At the current spatial location The fault signal transmitter was received. The signal strength that generates the fault signal Fireflies At the current spatial location The fault signal transmitter was received. The fault characteristic identification degree of the generated fault signals, Fireflies At the current spatial location The fault signal transmitter was received. The signal-to-noise ratio of the generated fault signal. Fireflies At the current spatial location The fault signal transmitter was received. The positioning accuracy of the generated fault signals. , , and They represent , , and The weighting coefficients.

[0021] Furthermore, in step S204, a distance-based penalty function is used to apply the penalty to the fireflies. The fitness is modified, including:

[0022] Traverse all fireflies in the current firefly population Fireflies with higher adaptability Calculating fireflies and fireflies Spatial distance between ;

[0023] Response to spatial distance Less than the preset penalty threshold For fireflies fitness Decrease the fitness to obtain the corrected fitness. , Represents the penalty function. This represents the penalty coefficient.

[0024] Further, in step S205, updating the spatial location of the fireflies based on the elite strategy includes:

[0025] Set up a Pareto archive to store elite solutions during the iteration process, including non-dominated solutions or excellent solutions representing different sensor layout schemes;

[0026] After the spatial location of fireflies is updated, several elite solutions are randomly selected from the Pareto archive periodically to replace the firefly individual with the worst fitness in the current firefly population, thereby updating the spatial location of fireflies.

[0027] Furthermore, in step S205, the engineering constraints include placement area constraints and electromagnetic compatibility constraints;

[0028] Regarding placement area constraints, if the candidate spatial location of the firefly does not meet the engineering constraints, the firefly will be placed... The spatial location is updated to the nearest feasible location, represented as: ;in, Indicates the updated spatial location. Indicates feasible regions and The difference set, This represents the set of geometric primitives corresponding to the non-placeable regions.

[0029] Further, step S300 includes the following sub-steps:

[0030] S301. Perform spatial cluster analysis on the sensor locations in each optimal sensor layout scheme;

[0031] S302. Perform overlap analysis on the spatial clustering analysis results to determine the locations of sensors that appear in each optimal sensor layout scheme or whose location distance is less than the set value, and mark them as the common optimal sensors for detecting multiple faults.

[0032] S303. In each optimal sensor arrangement scheme, identify the unique sensor position, mark it, and identify the characteristic sensor corresponding to the fault mode.

[0033] S304. By combining the optimal arrangement schemes of each sensor and marking the common optimal sensor and feature recognition sensor, the globally optimal sensor arrangement scheme is obtained.

[0034] Furthermore, step S200, the optimization process of the sensor layout scheme under each fault mode, includes a first stage and a second stage:

[0035] Phase 1: Implement the improved firefly algorithm to optimize the sensor layout scheme when collecting fault signals, and use Pareto archive to identify several distinct sensor layout scheme clusters;

[0036] The second stage involves executing a local search algorithm to fine-tune the sensor positions for the sensor layout cluster, thereby determining the optimal sensor layout scheme for each fault mode.

[0037] A sensor optimization arrangement system for large oil-filled electrical equipment, comprising:

[0038] Fault simulation module: Used to build a three-dimensional model of large oil-filled electrical equipment, determine the corresponding fault signal emission source by inputting the fault mode, and simulate the fault signal;

[0039] Sensor optimization module: Used to optimize sensor layout when sensors receive fault signals by executing an improved firefly algorithm;

[0040] Data storage module: used to store sensor layout schemes under various fault modes;

[0041] Comprehensive Analysis Module: Used to perform comprehensive analysis and processing of the optimal sensor layout scheme under various fault modes, and obtain the globally optimal sensor layout scheme with marked common optimal sensors and feature recognition sensors.

[0042] Furthermore, the sensor optimization module includes:

[0043] Fitness calculation unit: used to calculate and correct the fitness of fireflies using a weighted multi-factor objective function and a penalty function;

[0044] Spatial location optimization unit: used to update the spatial location of fireflies based on fitness, combined with elite strategies and engineering constraints.

[0045] The beneficial effects of this invention are as follows:

[0046] 1. Significantly improves the targeting of fault monitoring and the sensitivity of signal acquisition (based on reverse optimization of fault sources).

[0047] This invention establishes and optimizes signal source models for various specific fault modes. Existing technologies typically employ uniform or random arrangements based on geometric coverage, neglecting the complex internal structure of large oil-filled equipment and the significant differences in the locations of different fault sources. This invention abandons the common "wide coverage" strategy and instead establishes physical signal emission source models for specific typical faults (such as winding partial discharge and core overheating), using the firefly algorithm to reverse-engineer the optimal receiving location. This allows sensors to be precisely deployed at locations with minimal signal propagation attenuation and the most direct path, significantly improving the sensitivity to weak fault signals and ensuring high-intensity monitoring signals are obtained in the early stages of a fault.

[0048] 2. Significantly improved the quality of monitoring data and the accuracy of fault diagnosis (based on a weighted multi-factor objective function).

[0049] The improved firefly algorithm introduces a weighted multi-factor fitness function that includes signal strength, fault feature identification, signal-to-noise ratio, and positioning accuracy. Traditional wireless sensor network optimization only focuses on connectivity or coverage area, often resulting in sensors receiving signals but with severe signal distortion or extremely high noise. This invention constructs a fitness function that includes signal strength (… ), signal-to-noise ratio ( ), fault feature identification ( ) and positioning accuracy ( The multidimensional objective function of the sensor forces the algorithm to automatically weigh various factors during the optimization process. This mechanism ensures that the optimized sensor not only "can hear," but also "can hear clearly" and "preserves waveform fidelity," providing high-quality raw data for subsequent signal processing and fault mode identification, thereby effectively reducing the false alarm rate and the missed alarm rate.

[0050] 3. It overcomes the shortcomings of traditional algorithms that are prone to getting trapped in local optima and achieves comprehensive coverage of multi-mode faults (based on a multipolar optimization strategy).

[0051] The standard firefly algorithm, by introducing a distance-based penalty function, Pareto archiving, and an elite strategy, exhibits a strong tendency to cluster towards the current optimal solution, easily falling into a single local optimum trap. This leads to a large number of sensors congregating around a single strong signal point, while overlooking other important monitoring blind spots. This invention, by introducing a distance-based penalty function and a Pareto archiving mechanism, forces the population to maintain diversity (multipolarity). When sensors become overly clustered, their fitness is automatically reduced, guiding the fireflies to explore and lock onto multiple non-adjacent, excellent spatial layouts—the "polar points." This improvement, at the algorithmic level, ensures that the final solution can simultaneously cover multiple key monitoring areas, avoiding monitoring blind spots.

[0052] 4. It possesses extremely high engineering practicality and anti-interference capability (based on quantitative engineering constraint processing).

[0053] This invention mathematically encapsulates and integrates the rigid constraints of non-placeable areas and the flexible constraints of electromagnetic compatibility (EMC) into the algorithm. Instead of optimizing solely within an ideal mathematical model, it mathematically quantifies the actual physical constraints of large oil-filled equipment (such as radiators and valve obstructions) and the electromagnetic environment (such as strong EMI interference near high-pressure bushings). In particular, by incorporating EMC constraints as a noise term into the signal-to-noise ratio objective function, the algorithm can automatically avoid areas of strong interference or find the optimal balance between strong interference and strong signals. This allows the generated layout scheme to be implemented directly without manual adjustments, solving the problem of theoretical simulation results being unworkable in practical engineering.

[0054] 5. Achieved optimal allocation of global resources and accurate identification of specific faults (based on comprehensive analysis and feature sensor recognition).

[0055] Clustering and Feature Sensor Identification. Through comprehensive analysis of a single fault dataset, this invention can generate a globally optimal solution that considers all fault types, avoiding redundant sensor deployment and achieving the best balance between cost and performance. More importantly, this invention can identify the "feature recognition sensor" that responds most strongly to a specific fault. In actual operation, maintenance personnel only need to focus on the alarms of specific sensors to quickly deduce the fault type and approximate location, greatly shortening fault diagnosis time and improving equipment maintenance efficiency. Attached Figure Description

[0056] Figure 1 A flowchart illustrating the sensor optimization arrangement method for large oil-filled electrical equipment provided by this invention.

[0057] Figure 2 This is a schematic diagram illustrating the application of fault sources and sensor layout in a large oil-filled equipment provided by the present invention.

[0058] Figure 3 This is a schematic diagram comparing the multipolarity optimization effects provided by the present invention; wherein, Figure 3 (a) is a schematic diagram illustrating the effect of the traditional algorithm. Figure 3 (b) is a schematic diagram illustrating the effect of the improved algorithm of the present invention. Detailed Implementation

[0059] The specific embodiments of the present invention are described below to enable those skilled in the art to understand the present invention. However, it should be understood that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the present invention as defined and determined by the appended claims. All inventions utilizing the concept of the present invention are protected.

[0060] Example 1:

[0061] This invention provides a method for optimizing the arrangement of sensors in large oil-filled electrical equipment, such as... Figure 1 As shown, it includes the following steps:

[0062] S100. Establish a three-dimensional model of large oil-filled electrical equipment and set the fault signal emission source under various fault modes;

[0063] S200: Simulate the fault signals in the corresponding fault modes sequentially through the fault signal transmitter, and optimize the sensor arrangement scheme when collecting fault signals by executing the improved firefly algorithm to obtain the optimal sensor arrangement scheme for each fault mode.

[0064] In the improved firefly algorithm: the fitness of the firefly is calculated and corrected by a weighted multi-factor objective function and a penalty function. When updating the spatial position of the firefly based on the fitness, the updated spatial position of the firefly is determined by combining the elite strategy and engineering constraints.

[0065] S300: The optimal sensor layout scheme under various fault modes is comprehensively analyzed and processed to obtain the globally optimal sensor layout scheme with marked common optimal sensors and feature recognition sensors.

[0066] In step S100 of this embodiment, a three-dimensional model of a large oil-filled electrical device is established. Then, based on historical data and fault mechanism analysis, several typical fault modes that need to be monitored are set, and corresponding fault signal sources are set in the three-dimensional model for these fault modes. , This indicates the total number of failure modes.

[0067] For example, taking a large transformer as the object, a corresponding three-dimensional model is established, and the set fault modes include fault 1 (partial discharge at a certain point in the high-voltage winding), fault 2 (partial overheating at a certain point in the core), fault 3 (partial discharge at another point in the low-voltage winding), etc.

[0068] In step S200 of this embodiment, for any fault mode, the improved firefly algorithm is executed to optimize the sensor arrangement scheme when collecting fault signals, including:

[0069] S201. Based on the fault signal emission source corresponding to the fault mode, simulate the fault signal when the fault mode is generated.

[0070] S202. Within the feasible area around the 3D model, a firefly population is randomly generated as the initial sensor layout scheme, and each firefly corresponds to a sensor node and has a unique spatial location.

[0071] S203. Collect the fault signal generated by the fault signal transmitter received by each firefly at its current spatial location;

[0072] S204. Based on the fault signals received by each firefly, calculate and correct its fitness using a weighted multi-factor objective function.

[0073] S205. Based on the fitness of each firefly, determine the candidate spatial locations for firefly updates, and combine elite strategies and engineering constraints to determine the updated spatial locations of fireflies.

[0074] S206. Repeat steps S203 to S205 until the termination condition is met, and output the optimal sensor layout scheme corresponding to the current fault mode.

[0075] Based on the above process, for the first A fault mode is identified, and the corresponding fault signal emission source in the 3D model is activated to simulate the fault signal (such as ultrasonic signal or thermal signal) generated when the fault occurs.

[0076] Within a feasible area surrounding the 3D model (such as the surface of a transformer enclosure), a firefly population is randomly generated. Population size is Each firefly This represents a sensor node, whose spatial location is... .

[0077] In step S204, the fitness of fireflies... It is not a single coverage rate, but a weighted multi-factor function whose goal is to maximize the effectiveness of monitoring, expressed as:

[0078] ;

[0079] In the formula, Fireflies Adaptability, Fireflies At the current spatial location The fault signal transmitter was received. The signal strength that generates the fault signal Fireflies At the current spatial location The fault signal transmitter was received. The fault characteristic identification degree of the generated fault signals, Fireflies At the current spatial location The fault signal transmitter was received. The signal-to-noise ratio of the generated fault signal. Fireflies At the current spatial location The fault signal transmitter was received. The positioning accuracy of the generated fault signals. , , and They represent , , and The weighting coefficients.

[0080] Signal strength is used to quantify the signal energy received by the sensor. The signal attenuates as it propagates through the medium, which can be represented using a physics-based attenuation model.

[0081] ;

[0082] In the formula, It is a fault signal transmitter The initial signal amplitude; It is the Euclidean distance between the sensor and the source of the fault. It is the attenuation coefficient of a signal (such as ultrasonic or thermal signals) in a medium (such as transformer oil or steel plate); It is the geometric diffusion factor (e.g., for spherical diffusion). =2, during cylindrical diffusion =1).

[0083] Signal-to-noise ratio (SNR) measures the clarity of a signal in a noisy environment, and it is expressed as:

[0084] ;

[0085] in, It is a sensor Received from The signal power is usually related to Proportional It is a sensor The noise power at the location can be a composite function, for example... ,in It is uniformly distributed background noise, while It is location-dependent environmental noise (such as noise from sources of interference such as coolers and fans).

[0086] Fault feature identification is used to measure the received signal Compared with theoretical fault source signals Waveform similarity is used to combat signal distortion caused by multipath propagation, reflection, and dispersion. Normalized cross-correlation can be used as a metric.

[0087] ;

[0088] In the formula, It is a fault signal transmitter The theoretical characteristic waveform (e.g., a standard partial discharge pulse waveform). It is a sensor The actual received waveform. This value is between -1 and 1. The closer it is to 1, the less waveform distortion there is and the easier it is to identify the features.

[0089] Positioning accuracy is used to measure the position of the sensor. The contribution of accurate location to spatial fault source localization (e.g., via TDOA time difference of arrival) depends on the signal arrival time. For accurate estimation, the maximum rising edge slope of the signal is an effective indicator of timestamp accuracy. The steeper the slope (the less distortion), the more accurate the timing. ,in, It is the received signal waveform. A high slope means that the signal has a clear and sharp leading edge, which greatly improves the accuracy of the timestamp.

[0090] In the above formula, the weighting coefficient , , The sum is 1, which can be adjusted according to the specific monitoring task (e.g., whether to prioritize detection sensitivity or positioning accuracy).

[0091] It should be noted that the above formula is only an example, and other mathematical functions used to characterize signal strength, signal-to-noise ratio, fault feature identification, and positioning accuracy are also applicable.

[0092] In step S204, a distance-based penalty function is applied to the fireflies. The fitness is modified, including:

[0093] Traverse all fireflies in the current firefly population Fireflies with higher adaptability Calculating fireflies and fireflies Spatial distance between ;

[0094] Response to spatial distance Less than the preset penalty threshold For fireflies fitness Decrease the fitness to obtain the corrected fitness. , Represents the penalty function. This represents the penalty coefficient.

[0095] In this embodiment, a distance-based penalty function is introduced to modify the fitness, which reduces the fitness of explanations that are too close in space to the suboptimal point, thus preventing all fireflies from gathering at the same global optimum.

[0096] In step S205, the spatial location of the fireflies is updated based on the elite strategy, including:

[0097] During the iterative process of improving the firefly algorithm, a Pareto archive is set up to store elite solutions during the iteration process, including non-dominated solutions or excellent solutions representing different sensor layout schemes; for example, simultaneously considering... and The non-dominated solutions, or the distinct "extremes" found through clustering.

[0098] After the spatial location of fireflies is updated, several elite solutions are randomly selected from the Pareto archive periodically to replace the firefly individual with the worst fitness in the current firefly population, thereby updating the spatial location of fireflies.

[0099] Specifically, during the firefly location update process, the fireflies... Adaptability (brightness) Greater than fireflies Attract, move toward it. After updating the position, periodically (e.g., every...) (Generation) Randomly selects several "elite" solutions from the Pareto archive and replaces the current population. The firefly individual with the worst fitness. This can guide the population to continuously explore multiple discovered "extremes" and prevent the loss of an important layout scheme due to random drift.

[0100] In step S205, the position is updated based on the elite strategy. The inspection was conducted to determine whether it met engineering constraints, including placement area constraints and electromagnetic compatibility constraints. Specifically:

[0101] Regarding the placement area constraint, if the candidate spatial location of the firefly does not meet the placement area constraint, the firefly will be placed... The spatial location is updated to the nearest feasible location, represented as: ;in, Indicates the updated spatial location. Indicates feasible regions and The difference set, This represents the set of geometric primitives corresponding to the non-placeable regions.

[0102] The placement area constraint is a geometric constraint, and its quantification method is as follows:

[0103] First, all non-placeable areas (such as radiators, valves, supporting structures, etc.) are modeled in the 3D model as... A set of geometric primitives For regular shapes such as heat sinks or junction boxes, a cube primitive can be defined. The primitive is defined by its minimum coordinates. and maximum coordinates Limitations. For valves or pipes, a cylindrical primitive can be defined. This element is defined by its central axis. (Defined by a point and a direction vector) and radius limited.

[0104] Construct constraint judgment function A candidate position Deemed infeasible If it falls inside any geometric primitive: .

[0105] For example, for a cube , The judgment condition is: For example, for a cylinder , The judgment condition is: ,in The perpendicular distance from the point to the line.

[0106] When performing constraint processing, if Then the position The request was rejected. At this point, the "move to the nearest feasible location" operation can be performed. ,in Is the feasible area (e.g., the surface of the equipment) and The difference set. In the algorithm implementation, a simpler way is to directly assign the difference set to the firefly. Assign a very low fitness (e.g.) (or a very large negative value), causing it to be naturally eliminated or moved away from the region in the next iteration.

[0107] Electromagnetic compatibility (EMI) constraints are usually treated as flexible constraints during quantization because they are not a binary problem of "feasible / infeasible" but rather a matter of degree of "good / bad" (strong electromagnetic interference (EMI) reduces the signal-to-noise ratio (SNR) of a sensor).

[0108] In this embodiment, electromagnetic compatibility constraints are directly incorporated into the signal-to-noise ratio. In the objective function, noise power is quantized. To achieve this, noise power is defined as:

[0109] ;

[0110] Among them The noise power from the EMC interference source can be modeled as The modeling process is as follows:

[0111] Calibration in a 3D model The main sources of EMI (such as high-voltage bushings, high-current leads, core joints, etc.) are collectively referred to as follows: Each EMI source Electromagnetic fields (or their disturbance potentials) generated in space. With distance Attenuation can be modeled as:

[0112] ;

[0113] in, It is the first The strength of each interference source (e.g., determined by its voltage level or current amplitude). It is a sensor With interference source The distance between them; It is the decay exponent (e.g.) or ), total interference field It is all The linear superposition of sources is expressed as:

[0114] ;

[0115] Finally, the sensor Noise power induced by EMI With total interference field Proportional (or proportional to its square, depending on the coupling mechanism): in It is the electromagnetic coupling coefficient.

[0116] When performing constraint processing, Substitution Electromagnetic compatibility constraints are automatically and quantitatively included in the fitness function. of In this process, the spatial location of the firefly that satisfies the engineering constraints is obtained.

[0117] In step S206, the termination condition is reaching the maximum number of iterations or the fitness no longer significantly improves.

[0118] Step S300 of this implementation includes the following sub-steps:

[0119] S301. Perform spatial cluster analysis on the sensor locations in each optimal sensor layout scheme;

[0120] For example, All sensor locations in an optimal sensor layout scheme (e.g., scheme) Given the positions {1, 2, 3} in the given list, the solution is... Given the positions {4, 5, 6}, the solution is... The positions {7, 8, 9} in the middle, the solution Spatial clustering is performed on the positions {10, 11, 12} in the dataset;

[0121] S302. Perform overlap analysis on the spatial clustering analysis results to determine the locations of sensors that appear in each optimal sensor layout scheme or whose location distance is less than the set value, and mark them as the common optimal sensors for detecting multiple faults.

[0122] For example, positions 1 and 2 in the output;

[0123] S303. In each optimal sensor arrangement scheme, identify the unique sensor position, mark it, and identify the characteristic sensor corresponding to the fault mode.

[0124] Among them, the unique sensor location refers to a location that is only found in a specific scheme. In China (or against) The most unique position of the signal response, for example, 3, 5, 8, 11 in the output;

[0125] S304. Combine the optimal arrangement schemes of each sensor, and mark the common optimal sensor and feature recognition sensor to obtain the global optimal sensor arrangement scheme;

[0126] For example, the globally optimal sensor layout scheme This includes {positions 1, 2, 3, 5, 8, 11}, and explicitly marks {3} as the fault mode. The feature recognition sensor, {5} is the fault mode. The feature recognition sensor, {8} is the fault mode. The feature recognition sensor, {11} is the fault mode. Feature recognition sensor.

[0127] In step S200 of this embodiment, the optimization process of the sensor layout scheme under each fault mode includes a first stage and a second stage:

[0128] Phase 1: Implement the improved firefly algorithm to optimize the sensor layout scheme when collecting fault signals, and use Pareto archive to identify several distinct sensor layout scheme clusters;

[0129] The second stage involves executing a local search algorithm to fine-tune the sensor positions for the sensor layout cluster, thereby determining the optimal sensor layout scheme for each fault mode.

[0130] Specifically, the first phase involves a coarse search and "extreme" identification. In this phase, a large population and strong multipolarity penalty are used to quickly explore the entire feasible space and identify several distinct, promising clusters of sensor placement schemes (i.e., "extremes") using the Pareto archive. The second phase involves a fine-tuning search and placement adjustment. In this phase, a more localized and refined search is conducted around the "extremes" identified in phase one, fine-tuning the sensor positions to obtain the final optimal sensor placement scheme. For example, a smaller firefly population or other local search algorithms can be run again near the "extremes" selected in step S200.

[0131] Example 2:

[0132] This embodiment is a further limitation based on Embodiment 1. Its purpose is to provide a sensor optimization arrangement system for large oil-filled electrical equipment, which is implemented based on the sensor optimization arrangement method for large oil-filled electrical equipment in Embodiment 1. Other parts not mentioned refer to Embodiment 1 or the prior art.

[0133] This embodiment of a sensor optimization arrangement system for large oil-filled electrical equipment includes:

[0134] Fault simulation module: Used to build a three-dimensional model of large oil-filled electrical equipment, determine the corresponding fault signal emission source by inputting the fault mode, and simulate the fault signal;

[0135] Sensor optimization module: Used to optimize sensor layout when sensors receive fault signals by executing an improved firefly algorithm;

[0136] Data storage module: used to store sensor layout schemes under various fault modes;

[0137] Comprehensive Analysis Module: Used to perform comprehensive analysis and processing of the optimal sensor layout scheme under various fault modes, and obtain the globally optimal sensor layout scheme with marked common optimal sensors and feature recognition sensors.

[0138] In this embodiment, the sensor optimization module includes:

[0139] Fitness calculation unit: used to calculate and correct the fitness of fireflies using a weighted multi-factor objective function and a penalty function;

[0140] Spatial location optimization unit: used to update the spatial location of fireflies based on fitness, combined with elite strategies and engineering constraints.

[0141] Specifically, the weighted multi-factor objective function in the fitness calculation unit is expressed as:

[0142] ;

[0143] In the formula, Fireflies Adaptability, Fireflies At the current spatial location The fault signal transmitter was received. The signal strength that generates the fault signal Fireflies At the current spatial location The fault signal transmitter was received. The fault characteristic identification degree of the generated fault signals, Fireflies At the current spatial location The fault signal transmitter was received. The signal-to-noise ratio of the generated fault signal. Fireflies At the current spatial location The fault signal transmitter was received. The positioning accuracy of the generated fault signals. , , and They represent , , and The weighting coefficients.

[0144] In the fitness calculation unit, a distance-based penalty function is used to evaluate the fitness of fireflies. The fitness is modified, including:

[0145] Traverse all fireflies in the current firefly population Fireflies with higher adaptability Calculating fireflies and fireflies Spatial distance between ;

[0146] Response to spatial distance Less than the preset penalty threshold For fireflies fitness Decrease the fitness to obtain the corrected fitness. , Represents the penalty function. This represents the penalty coefficient.

[0147] In the spatial location optimization unit, the spatial location of fireflies is updated based on an elite strategy, including:

[0148] Set up a Pareto archive to store elite solutions during the iteration process, including non-dominated solutions or excellent solutions representing different sensor layout schemes;

[0149] After the spatial location of fireflies is updated, several elite solutions are randomly selected from the Pareto archive periodically to replace the firefly individual with the worst fitness in the current firefly population, thereby updating the spatial location of fireflies.

[0150] In the spatial location optimization unit, engineering constraints include placement area constraints and electromagnetic compatibility constraints; specifically, regarding placement area constraints, when a firefly's candidate spatial location does not meet the engineering constraints, the firefly is... The spatial location is updated to the nearest feasible location, represented as: ;in, Indicates the updated spatial location. Indicates feasible regions and The difference set, This represents the set of geometric primitives corresponding to the non-placeable regions; for electromagnetic compatibility constraints, the signal-to-noise ratio is directly incorporated into the weighted multi-factor objective function. In the process, noise power is quantized. This can be achieved through a specific item.

[0151] To more intuitively understand the application scenarios of this invention, please refer to [link / reference]. Figure 2 It shows a schematic diagram of the application of fault sources and sensor arrangement in a large oil-filled equipment in one embodiment of this application.

[0152] like Figure 2 As shown, large oil-filled electrical equipment (such as transformers) is a complex system where multiple faults may occur at different locations. Examples include fault source 1, fault source 2, and fault source a. Traditional sensor placement methods (the "initial sensors" marked with 'X' in the figure) may be based on engineering experience or a uniform distribution, resulting in a distance from the actual fault sources and weak signal acquisition capabilities. The optimized placement method of this invention, through the optimization process of the improved firefly algorithm described above, can find a set of better sensor locations (the "optimized sensors" marked with '●' in the figure) for these specific fault sources (step S100). These optimized locations are closer to the optimal propagation path of the fault signal, thereby significantly improving monitoring sensitivity and signal-to-noise ratio. Furthermore, the sensors marked with '▲' in the figure (e.g., sensors near fault source 1 and fault source a) are identified as feature recognition sensors, which have the strongest signal response to specific faults and can be used to accurately distinguish different fault types.

[0153] To further illustrate the superiority of the improved firefly algorithm of this invention compared to the prior art, please refer to [link / reference needed]. Figure 3The diagram illustrates a comparison of the effects of multipolar optimization. In this diagram, it is assumed that there are three potential optimal arrangement regions: "Pole S1," "Pole S2," and "Pole S3," each corresponding to a different failure mode or optimization direction.

[0154] See Figure 3 (a) shows the optimization effect of the traditional firefly algorithm. Due to the lack of a multi-polarity search mechanism, all fireflies (representing candidate sensor nodes) eventually cluster around only one strongest pole (e.g., "pole S1"), completely ignoring "pole S2" and "pole S3". This situation of being trapped in a single optimal solution causes the algorithm to only find the optimal solution for one type of fault when dealing with multiple fault modes, while losing the layout schemes for other faults.

[0155] See Figure 3 (b) illustrates the optimization effect of the improved algorithm of the present invention. As shown in the figure, the firefly population was successfully guided to all three poles ("Pole S1", "Pole S2", and "Pole S3") and formed clusters respectively. This shows that the method of the present invention can simultaneously explore and retain multiple distinct excellent layout schemes, providing a complete data foundation for subsequent comprehensive analysis and ensuring that the final globally optimal sensor layout scheme can take into account all fault types.

[0156] In summary, the sensor optimization layout method and system for large oil-filled electrical equipment based on the improved firefly algorithm provided by this invention overcomes the shortcomings of traditional methods, such as reliance on experience, single objective, and low measurement accuracy, by introducing successive optimization for multiple fault modes, multi-polarity improvement of the algorithm, and a final comprehensive analysis process. This invention can provide a high-precision, high-fault-identification sensor layout scheme according to the specific structure and typical fault characteristics of the equipment, which not only improves the safety and reliability of equipment operation but also provides crucial data support for subsequent condition assessment and fault diagnosis.

[0157] Specific embodiments have been used to illustrate the principles and implementation methods of this invention. The descriptions of the embodiments above are only for the purpose of helping to understand the method and core ideas of this invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this invention. Therefore, the content of this specification should not be construed as a limitation of this invention.

[0158] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the principles of the invention, and should be understood that the scope of protection of the invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of this invention.

Claims

1. A method for optimizing the arrangement of sensors in large oil-filled electrical equipment, characterized in that, Includes the following steps: S100. Establish a three-dimensional model of large oil-filled electrical equipment and set the fault signal emission source under various fault modes; S200: Simulate the fault signals in the corresponding fault modes sequentially through the fault signal transmitter, and optimize the sensor arrangement scheme when collecting fault signals by executing the improved firefly algorithm to obtain the optimal sensor arrangement scheme for each fault mode. In the improved firefly algorithm: the fitness of the firefly is calculated and corrected by a weighted multi-factor objective function and a penalty function. When updating the spatial position of the firefly based on the fitness, the updated spatial position of the firefly is determined by combining an elite strategy and engineering constraints. S300: The optimal sensor layout scheme under various fault modes is comprehensively analyzed and processed to obtain the globally optimal sensor layout scheme with marked common optimal sensors and feature recognition sensors.

2. The method for optimizing the arrangement of sensors in large oil-filled electrical equipment according to claim 1, characterized in that, In step S200, for any fault mode, the improved firefly algorithm is executed to optimize the sensor arrangement scheme when collecting fault signals, including: S201. Based on the fault signal emission source corresponding to the fault mode, simulate the fault signal when the fault mode is generated. S202. Within the feasible area around the 3D model, a firefly population is randomly generated as the initial sensor layout scheme, and each firefly corresponds to a sensor node and has a unique spatial location. S203. Collect the fault signal generated by the fault signal transmitter received by each firefly at its current spatial location; S204. Based on the fault signals received by each firefly, calculate and correct its fitness using a weighted multi-factor objective function. S205. Based on the fitness of each firefly, determine the candidate spatial locations for firefly updates, and combine elite strategies and engineering constraints to determine the updated spatial locations of fireflies. S206. Repeat steps S203 to S205 until the termination condition is met, and output the optimal sensor layout scheme corresponding to the current fault mode.

3. The method for optimizing the arrangement of sensors in large oil-filled electrical equipment according to claim 2, characterized in that, In step S204, the weighted multi-factor objective function is expressed as follows: ； In the formula, Fireflies Adaptability, Fireflies At the current spatial location The fault signal transmitter was received. The signal strength that generates the fault signal Fireflies At the current spatial location The fault signal transmitter was received. The fault characteristic identification degree of the generated fault signals, Fireflies At the current spatial location The fault signal transmitter was received. The signal-to-noise ratio of the generated fault signal. Fireflies At the current spatial location The fault signal transmitter was received. The positioning accuracy of the generated fault signals. , , and They represent , , and The weighting coefficients.

4. The method for optimizing the arrangement of sensors in large oil-filled electrical equipment according to claim 2, characterized in that, In step S204, a distance-based penalty function is used to apply the penalty to the fireflies. The fitness is modified, including: Traverse all fireflies in the current firefly population Fireflies with higher adaptability Calculating fireflies and fireflies Spatial distance between ; Response to spatial distance Less than the preset penalty threshold For fireflies fitness Decrease the fitness to obtain the corrected fitness. , Represents the penalty function. This represents the penalty coefficient.

5. The method for optimizing the arrangement of sensors in large oil-filled electrical equipment according to claim 2, characterized in that, In step S205, updating the spatial location of fireflies based on an elite strategy includes: Set up a Pareto archive to store elite solutions during the iteration process, including non-dominated solutions or excellent solutions representing different sensor layout schemes; After the spatial location of fireflies is updated, several elite solutions are randomly selected from the Pareto archive periodically to replace the firefly individual with the worst fitness in the current firefly population, thereby updating the spatial location of fireflies.

6. The method for optimizing the arrangement of sensors in large oil-filled electrical equipment according to claim 2, characterized in that, In step S205, the engineering constraints include placement area constraints and electromagnetic compatibility constraints. Regarding placement area constraints, if the candidate spatial location of the firefly does not meet the engineering constraints, the firefly will be placed... The spatial location is updated to the nearest feasible location, represented as: ;in, Indicates the updated spatial location. Indicates feasible regions and The difference set, This represents the set of geometric primitives corresponding to the non-placeable regions.

7. The method for optimizing the arrangement of sensors in large oil-filled electrical equipment according to claim 1, characterized in that, Step S300 includes the following sub-steps: S301. Perform spatial cluster analysis on the sensor locations in each optimal sensor layout scheme; S302. Perform overlap analysis on the spatial clustering analysis results to determine the locations of sensors that appear in each optimal sensor layout scheme or whose location distance is less than the set value, and mark them as the common optimal sensors for detecting multiple faults. S303. In each optimal sensor arrangement scheme, identify the unique sensor position, mark it, and identify the characteristic sensor corresponding to the fault mode. S304. By combining the optimal arrangement schemes of each sensor and marking the common optimal sensor and feature recognition sensor, the globally optimal sensor arrangement scheme is obtained.

8. The method for optimizing the arrangement of sensors in oil-filled electrical equipment according to claim 1, characterized in that, Step S200, the optimization process of the sensor layout scheme under each fault mode, includes a first stage and a second stage: Phase 1: Implement the improved firefly algorithm to optimize the sensor layout scheme when collecting fault signals, and use Pareto archive to identify several distinct sensor layout scheme clusters; The second stage involves executing a local search algorithm to fine-tune the sensor positions for the sensor layout cluster, thereby determining the optimal sensor layout scheme for each fault mode.

9. A sensor optimization arrangement system for large oil-filled electrical equipment, implemented based on the sensor optimization arrangement method for large oil-filled electrical equipment according to any one of claims 1 to 8, characterized in that, include: Fault simulation module: Used to build a three-dimensional model of large oil-filled electrical equipment, determine the corresponding fault signal emission source by inputting the fault mode, and simulate the fault signal; Sensor optimization module: Used to optimize sensor layout when sensors receive fault signals by executing an improved firefly algorithm; Data storage module: used to store sensor layout schemes under various fault modes; Comprehensive Analysis Module: Used to perform comprehensive analysis and processing of the optimal sensor layout scheme under various fault modes, and obtain the globally optimal sensor layout scheme with marked common optimal sensors and feature recognition sensors.

10. The sensor optimization arrangement system for large oil-filled electrical equipment according to claim 9, characterized in that, The sensor optimization module includes: Fitness calculation unit: used to calculate and correct the fitness of fireflies using a weighted multi-factor objective function and a penalty function; Spatial location optimization unit: used to update the spatial location of fireflies based on fitness, combined with elite strategies and engineering constraints.

Images