A high-voltage cable passage lighting intelligent control method based on data analysis

Abstract

The application relates to the field of lighting control, in particular to a high-voltage cable channel lighting intelligent control method based on data analysis. The high-voltage cable channel lighting intelligent control method based on data analysis comprises the following steps: S1, acquiring equipment information of a high-voltage cable channel, and screening a reflective equipment sequence according to the equipment information; S2, processing historical operation and maintenance data of the high-voltage cable channel to obtain inspection target information and inspection position information; S3, acquiring lamp information of the high-voltage cable channel, constructing a lighting direct radiation area graph based on the lamp information, and constructing a direct radiation dead angle target position graph according to the lighting direct radiation area graph and the inspection target information; and S4, obtaining a reflection area position and a reflection range through a lamp light reflection model according to the lamp information and the reflective equipment sequence. The application can realize effective light compensation by utilizing the reflection characteristics of equipment in the channel, and can inhibit the interference of adverse reflection on detection sensors.

Description

Technical Field

[0001] This invention relates to the field of lighting control, and more particularly to a smart control method for lighting in high-voltage cable channels based on data analysis. Background Technology

[0002] In power systems, high-voltage cable channels are typically long, narrow, enclosed, and structurally complex underground or semi-underground spaces. Along the channel, a large number of cable bodies, supports, grounding devices, integrated online monitoring equipment, and auxiliary facilities are laid out. These devices are made of various materials and have different surface shapes. During manual or robotic inspections, high requirements are placed on the continuity, stability, and effective coverage of lighting. Especially in the presence of metal shells and coated components with high reflectivity, the light distribution in the channel often exhibits obvious unevenness. With the continuous expansion of the power grid and the improvement of intelligent operation and maintenance, high-voltage cable channels are gradually evolving from manual inspection to robotic inspection and comprehensive online monitoring. The inspection targets have evolved from simple cable appearance checks to refined detection of multi-dimensional conditions such as partial discharge, temperature, deformation, and leakage. This means that lighting systems not only need to meet basic channel visibility requirements, but also need to provide adapted illuminance environments for different inspection targets and sensor locations. In existing technologies, lighting methods for high-voltage cable channels mainly include fixed-brightness lighting, zoned timed lighting, and lighting control methods based on simple sensor triggers. For example, using infrared or human body sensors to achieve "lights on when someone is present, lights off when no one is present," or turning on the channel lighting as a whole or in sections according to a preset inspection route. Although the above-mentioned existing technologies reduce energy consumption and manual operation to a certain extent, most of them only focus on turning the lighting on and off or adjusting the overall brightness, without fully considering the light reflection effect caused by the complex equipment layout in the channel, nor do they combine historical operation and maintenance data and inspection plans to conduct refined analysis of lighting requirements. This leads to problems such as direct lighting blind spots, local overexposure, or reflected glare interference in practical applications. Therefore, there is an urgent need for a data analysis-based intelligent control method for lighting in high-voltage cable tunnels, which can effectively supplement lighting by utilizing the reflective characteristics of equipment within the tunnel, and suppress interference from adverse reflections on detection sensors, thereby improving the efficiency of lighting resource utilization and the safety level of cable tunnel operation and maintenance. Summary of the Invention

[0003] To overcome the shortcomings of existing technologies in accurately matching the needs of robot inspection, this invention provides a data analysis-based intelligent control method for lighting in high-voltage cable channels.

[0004] The technical implementation scheme of the present invention is: a smart control method for lighting in high-voltage cable tunnels based on data analysis, comprising the following steps: S1: Obtain equipment information for the high-voltage cable channel, and filter to obtain a sequence of reflective devices based on the equipment information; S2: Process historical operation and maintenance data of high-voltage cable channels to obtain inspection target information; S3: Obtain the lighting information of the high-voltage cable channel, construct a direct lighting area map based on the lighting information, and construct a direct lighting blind spot target location map based on the direct lighting area map and the inspection target information; S4: Based on the lamp information and the reflective device sequence, the location and range of the reflection area are obtained through the lamp light reflection model; S5: Obtain the inspection plan data of the inspection robot, and construct a reflection gain interference conversion sequence based on the inspection plan data, the location of the reflection area, and the reflection range; S6: Based on the lamp information, the target position map of the direct blind spot, and the reflective device sequence, solve the directional lamp adjustment sequence; based on the lamp information and the reflection gain interference conversion sequence, solve the non-directional lamp adjustment sequence; and perform lighting adjustment control based on the non-directional lamp adjustment sequence and the directional lamp adjustment sequence.

[0005] Preferably, the step of obtaining equipment information of the high-voltage cable channel and filtering to obtain a reflective equipment sequence based on the equipment information includes: the equipment information includes equipment identification, equipment type, equipment location, effective reflective area and surface reflectivity; the equipment is divided into reflective equipment and non-reflective equipment based on the surface reflectivity; and the equipment identification and surface reflectivity of the reflective equipment are used to construct a reflective equipment sequence.

[0006] Preferably, the step of processing historical operation and maintenance data of the high-voltage cable channel to obtain inspection target information includes: acquiring historical operation and maintenance data of the high-voltage cable channel, performing natural language processing on the historical operation and maintenance data to obtain inspection target information, wherein the inspection target information includes the required illumination of the inspection target and the location of the inspection target.

[0007] Preferably, the step of acquiring the lighting information of the high-voltage cable channel, constructing a direct illumination area map based on the lighting information, and constructing a direct illumination blind spot target location map based on the direct illumination area map and the inspection target information includes: the lighting information includes directional lighting positions, non-directional lighting positions, and lighting parameters; the high-voltage cable channel area is gridded; the illuminance of each grid is solved using a direct illumination model based on the non-directional lighting positions and lighting parameters to construct the direct illumination area map; the direct illumination area map is used as a reference to determine the direct illumination blind spot target location based on the inspection target location, thereby generating the direct illumination blind spot target location map.

[0008] Preferably, the step of solving the direct illumination area using a direct illuminance model based on the non-directional luminaire position and luminaire parameters includes: the direct illuminance model formula is: ; In the formula, Non-directional lighting fixtures In position direct illuminance, Non-directional lighting fixtures In the main axis and position of the lamp included angle Light intensity distribution, Non-directional lighting fixtures Arrive at the location The square of the distance, Non-directional lighting fixtures In position The cosine of the angle between the normal of the illuminated surface and the incident direction. Non-directional lighting fixtures The dimming ratio.

[0009] Preferably, obtaining the reflection area position and reflection range based on the lamp information and the reflector sequence using a lamp light reflection model includes: obtaining the reflector positions through the reflector sequence; and solving for the reflection area position and reflection range using a lamp light reflection model based on the directional lamp positions, non-directional lamp positions, lamp parameters, and reflector positions, wherein the reflection area position is taken as the center point position of the reflection area. The formula for the light reflection model of a lamp is: ; In the formula, For lighting fixtures Through the device In position Illuminance of reflected light, For lighting fixtures dimming ratio, For equipment Surface reflectivity, For lighting fixtures In the equipment Incident direction and equipment The angle between the normals, For equipment Arrive at the location The square of the distance, For equipment The effective reflective area, To pass through the device The reflected light at position The cosine of the angle between the normal of the surface receiving light and the incident direction; ; In the formula, For lighting fixtures Through the device In position The radius of the resulting reflection range, For equipment Arrive at the location distance, The equivalent divergence half-angle of the lamp The tangent value, This is the scaling factor from dimension to radius; Preferably, the step of acquiring the inspection plan data of the inspection robot and constructing a reflection gain-interference conversion sequence based on the inspection plan data, the reflection area location, and the reflection range includes: acquiring the inspection plan data of the inspection robot; processing the inspection plan data using natural language processing technology to obtain planned inspection target information and planned inspection location information; acquiring the planned inspection target location based on the planned inspection target information; acquiring the planned detection sensor location based on the planned inspection location information; comparing the reflection range with the planned detection sensor location and the planned inspection target location; if the planned detection sensor location is within the reflection range, calculating the gain-interference conversion coefficient; marking the gain or interference state based on the gain-interference conversion coefficient; if the planned detection sensor location is not within the reflection range but the planned inspection target location is within the reflection range, marking it as a gain state; and generating a reflection gain-interference conversion sequence based on the marking results.

[0010] Preferably, if the planned sensor is located within the reflection range, calculating the gain-interference conversion coefficient and marking the gain or interference state based on the gain-interference conversion coefficient includes: The formula for the gain-interference conversion factor is: ; In the formula, For gain-to-interference conversion coefficient, The gain intensity brought about by reflection Interference intensity; ; In the formula, For reflected illuminance, To detect the safe illuminance threshold of the sensor; ; In the formula, For direct illuminance, It is a very small positive number.

[0011] Preferably, the step of solving the directional lighting adjustment sequence based on the lighting fixture information, the direct blind spot target position map, and the reflector sequence, and solving the non-directional lighting adjustment sequence based on the lighting fixture information and the reflection gain interference conversion sequence, and performing lighting adjustment control in combination with the directional lighting adjustment sequence, includes: obtaining the required illuminance of the planned inspection blind spot target position based on the planned inspection target position, and optimizing the directional lighting adjustment sequence in combination with the directional lighting fixture position, the reflector position, and the direct blind spot target position map; generating the reflection gain interference conversion sequence according to the planned inspection target position, and optimizing the non-directional lighting adjustment sequence in combination with the non-directional lighting fixture position; and sending precise adjustment commands, including azimuth adjustment, pitch adjustment, and dimming ratio adjustment, to each directional lighting fixture through the lighting fixture control unit according to the directional lighting adjustment sequence, and sending a dimming ratio adjustment command to each non-directional lighting fixture according to the non-directional lighting adjustment sequence.

[0012] Preferably, the step of optimizing the directional lighting fixture adjustment sequence by combining the location of the directional lighting fixture, the location of the reflector, and the target location map of the direct blind spot; generating a reflection gain interference conversion sequence based on the planned inspection target location, and optimizing the non-directional lighting fixture adjustment sequence by combining the location of the non-directional lighting fixture, includes: Formula for the objective function of directional lighting optimization: ; In the formula, To optimize the objective function value for directional lighting fixtures, Blind spot target location The required illuminance, For the number of directional lights, For directional lighting fixtures Target location in the blind spot Illuminance of reflected light, For directional lighting fixtures The square of the azimuth adjustment amount, For directional lighting fixtures The square of the pitch angle adjustment amount, For directional lighting fixtures The square of the dimming ratio adjustment amount. and These are the weighting coefficients; Formula for the objective function of non-directional lighting fixture optimization: ; In the formula, Optimize the objective function value for non-directional luminaires. For the inspection target location The required illuminance, For the inspection target location The sum of direct illuminance and reflected illuminance, Non-directional lighting fixtures The square of the dimming ratio adjustment amount. At the inspection target location Gain-interference conversion coefficient at the location, , and These are the weighting coefficients.

[0013] The present invention has the following advantages: This invention integrates historical maintenance data with real-time equipment information to construct a refined lighting model that includes both direct and reflected light paths. This model proactively identifies and distinguishes between blind spots requiring enhanced illumination and interference-causing reflective areas. On one hand, directional lighting fixtures are adjusted to provide targeted and precise illumination to blind spots using reflective devices, ensuring that the inspected targets reach the required illuminance. On the other hand, the invention proactively predicts and quantifies the interference risk of reflected light on the inspection robot's sensors. By optimizing non-directional lighting fixture parameters or adjusting the lighting strategy associated with the inspection plan, harmful glare is effectively suppressed, ensuring the stability and accuracy of optical sensor data acquisition and improving the detection rate and reliability of automated inspections.

[0014] This invention elevates lighting control from an independent subsystem to an intelligent node deeply integrated with the equipment management system and inspection robot system. It can automatically process maintenance records and inspection plans described in natural language, understand inspection intentions and spatial requirements, and dynamically generate lighting configurations accordingly. This adaptive capability allows lighting strategies to automatically adjust to changes in inspection tasks, equipment layout, and corridor structure modifications, eliminating the need for manual reprogramming or scene setting. This significantly reduces the complexity and labor costs of long-term maintenance, and enhances the intelligence and integration of the entire corridor maintenance management system. Attached Figure Description

[0015] Figure 1 This is a flowchart of the intelligent control method for high-voltage cable channel lighting based on data analysis according to the present invention.

[0016] Figure 2 This is a flowchart illustrating the intelligent control decision-making process for high-voltage cable channel lighting based on data analysis, as described in this invention. Detailed Implementation

[0017] The invention will now be described more fully below with reference to the accompanying drawings, in which presently preferred embodiments of the invention are illustrated. However, the invention may be embodied in many different forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided for thoroughness and completeness and to fully convey the scope of the invention to those skilled in the art.

[0018] A data analysis-based intelligent control method for lighting in high-voltage cable tunnels, such as... Figure 1 and Figure 2 As shown, it includes the following steps: S1: Obtain equipment information for the high-voltage cable channel, and filter to obtain a sequence of reflective devices based on the equipment information; The equipment information includes equipment identification, equipment type, equipment location, effective reflective area, and surface reflectivity. Based on the surface reflectivity, the equipment is divided into reflective equipment and non-reflective equipment. The equipment identification and surface reflectivity of the reflective equipment are obtained to construct a reflective equipment sequence.

[0019] It should be further explained that, during the implementation phase, step S1 first extracts complete information on all relevant equipment within the high-voltage cable channel from the existing equipment database. This equipment information exists in the form of structured data, covering several key data fields, including equipment identifiers designed to uniquely distinguish different devices, equipment types describing the functional affiliation or specific category of the equipment, equipment positions precisely calibrated in the three-dimensional space of the channel, effective reflective area used to quantify the physical area of ​​the equipment surface that participates in light reflection, and surface reflectivity, a physical quantity characterizing the level of light reflection capability of the equipment surface.

[0020] After comprehensively collecting the aforementioned equipment information, all equipment is automatically classified based on surface reflectivity, a core optical characteristic. A surface reflectivity threshold is pre-set, determined based on the reflectivity characteristics of typical equipment materials within the channel and the effectiveness of supplementary lighting. Typical surface reflectivity values ​​for common equipment materials (such as metals, ceramics, and paint coatings) within the channel are sampled and measured using standard optical materials handbooks or portable reflectivity meters to obtain their distribution range. Secondly, combining lighting simulation and experiments, the reflection efficiency of different reflectivity surfaces for incident light from directional luminaires at typical distances and angles, and their contribution to supplementary lighting in target blind spots, are analyzed. The analysis reveals that when the surface reflectivity is below a certain critical value, the reflected light intensity is too low to provide effective supplementary illumination for blind spots (i.e., the reflected illuminance value is lower than the minimum incremental illuminance required for effective detection by the inspection sensor, for example, 5 lux). Based on this, the minimum reflectivity threshold that can provide effective supplementary lighting is set as the surface reflectivity threshold, and the measured surface reflectivity value of each device is compared with this threshold one by one. Based on the comparison results, the equipment was divided into two distinct categories: reflective equipment and non-reflective equipment. Equipment with a surface reflectivity greater than or equal to a set threshold was classified as reflective equipment, which is commonly found in devices with smooth surfaces such as metal casings or ceramic insulators. Conversely, equipment with a surface reflectivity lower than the set threshold was classified as non-reflective equipment, such as cable sheath materials with rough surfaces or dark colors.

[0021] After the classification process is completed, key data is extracted from the identified reflective device categories. The device identifier and its surface reflectivity value associated with each reflective device entry are read sequentially, and this pair of information is combined into a single associated data unit and saved. Next, following a preset data organization format, such as a list or array, the data units corresponding to all reflective devices in the channel are arranged and integrated in an orderly manner, constructing a clearly structured and ordered sequence of reflective devices. This sequence of reflective devices constitutes the key data input for subsequent establishment of light reflection models and analysis of reflection effects; each record in the sequence uniquely corresponds to an actual physical device with specific reflective properties in the channel.

[0022] S2: Process historical operation and maintenance data of high-voltage cable channels to obtain inspection target information; Historical operation and maintenance data of high-voltage cable channels are acquired, and natural language processing is performed on the historical operation and maintenance data to obtain inspection target information, which includes the required illumination of the inspection target and the location of the inspection target.

[0023] It should be further explained that in the implementation of step S2, historical maintenance data accumulated over a long period of time is retrieved from the relevant operation and maintenance management platform or a dedicated historical database for the high-voltage cable channel. This data is mostly in unstructured text format, commonly found in work order records filled out by maintenance personnel, periodically generated inspection reports, detailed fault description logs, and daily work handover notes. The text content essentially carries key information about the inspection status of equipment within the channel, the process of various maintenance activities, and the handling of abnormal problems. Natural language processing is used to automatically parse and process the acquired historical text data. The core purpose of this processing is to accurately identify and extract two types of structured information from the text description that directly influence subsequent intelligent lighting decisions: inspection target information and inspection location information.

[0024] The natural language processing (NLP) process comprehensively utilizes various text analysis techniques. Through rule-based pattern matching and named entity recognition (NER) technologies, it extracts equipment names, serial numbers, and associated directional descriptive terms (such as "east" and "above") mentioned in the text. The required illuminance value for the inspection target is obtained by matching keywords in the text (such as "fine inspection" and "infrared temperature measurement") with a pre-defined "Inspection Task-Illuminance Requirement Mapping Table," which is predefined by those skilled in the art based on sensor characteristics and industry standards. For example, the keyword "partial discharge detection" maps to 300 lux, and "visual inspection" maps to 150 lux. The location of the inspection target is obtained by querying pre-stored three-dimensional coordinates in the equipment asset database using the parsed equipment identifier. After the above processing, the output inspection target information and inspection location information, which have been converted into a structured format, are uniformly standardized and packaged. This refined information collectively completes the digital reconstruction of historical operation and maintenance experience and work habits, thus laying a data foundation directly based on actual operational history for subsequent stages of building accurate lighting demand models and implementing forward-looking lighting strategy configurations.

[0025] S3: Obtain the lighting information of the high-voltage cable channel, construct a direct lighting area map based on the lighting information, and construct a direct lighting blind spot target location map based on the direct lighting area map and the inspection target information; The lighting information includes the location of directional lighting fixtures, the location of non-directional lighting fixtures, and lighting fixture parameters. The high-voltage cable channel area is gridded. Based on the location of the non-directional lighting fixtures and the lighting fixture parameters, the illuminance of each grid is solved using a direct illuminance model to construct a direct lighting area map. Using the direct lighting area map as a reference and the location of the inspection target, the location of the direct lighting blind spot target is determined, and a direct lighting blind spot target location map is generated.

[0026] The direct illumination area is determined using a direct illuminance model based on the non-directional luminaire positions and luminaire parameters. The direct illuminance model formula is as follows: ; In the formula, Non-directional lighting fixtures In position direct illuminance, Non-directional lighting fixtures In the main axis and position of the lamp included angle Light intensity distribution, Non-directional lighting fixtures Arrive at the location The square of the distance is for non-directional lighting fixtures. Arrive at the location distance, Non-directional lighting fixtures In position The cosine of the angle between the normal of the illuminated surface and the incident direction. Non-directional lighting fixtures The dimming ratio.

[0027] It should be further explained that the specific implementation process of step S3 is as follows: Complete luminaire information is obtained from the lighting management system or relevant database of the high-voltage cable channel. The luminaire information is a structured data set that explicitly includes the following: the installation positions of all directional luminaires capable of adjusting their illumination direction within the channel, i.e., directional luminaire positions; the installation positions of all non-directional luminaires providing floodlight illumination with a fixed illumination direction within the channel, i.e., non-directional luminaire positions; various parameters describing the optical and electrical characteristics of the luminaires, i.e., luminaire parameters, such as rated luminous flux, light distribution curve, dimming ratio, and power; and a unique number or code used to identify each luminaire, i.e., luminaire identification.

[0028] After obtaining the luminaire information, the entire three-dimensional space covered by the high-voltage cable channel is discretized, i.e., the high-voltage cable channel area is meshed. This meshing process divides the continuous channel space into a large number of small, regular cubic units, each representing an independently calculated and analyzed spatial point. Based on the location and parameters of the non-directional luminaires, the direct illumination area of ​​the entire channel space is solved using a direct illuminance model. Specifically, for any point in the space, the direct illuminance contribution of each non-directional luminaire to that point is calculated using the direct illuminance model. The direct illuminance model formula is used to calculate the illuminance value of direct illumination produced by a certain non-directional luminaire at any point within the high-voltage cable channel, where... Non-directional lighting fixtures In position Direct illuminance, measured in lux; This represents the angle between the luminaire's optical axis and the ray pointing to its position. The luminous intensity distribution in a given direction is determined by the light distribution curve of the luminaire, and the unit is candela. It is a lighting fixture With position The square of the straight-line distance between them reflects the inverse square law of light attenuation with distance; It is a location The normal direction of the illuminated plane and the direction from the luminaire The cosine of the angle between the incident light direction and the incident light direction represents the influence of the incident light angle on the illumination. When the angle is greater than 90 degrees (the cosine value is negative), it means that the point is located on the back surface of the lamp. The maximum value function is used to set it to zero to eliminate invalid illumination. This is the current dimming ratio of the non-directional luminaire, a dimensionless number between 0 and 1, used to represent the ratio of the luminaire's actual output luminous flux to its maximum rated value. Dynamic control of illuminance is achieved by adjusting this coefficient.

[0029] The above model is applied to each grid point to accumulate the direct illuminance contribution of all non-directional luminaires at that point, thereby obtaining detailed illuminance distribution data generated by non-directional luminaires within the channel. Based on this distribution data, the system uses a preset minimum illuminance threshold that meets basic visibility requirements as a boundary, and determines the continuous spatial area formed by all grid points with illuminance values ​​higher than this threshold as the effectively illuminated area, thus constructing a direct illumination area map that clearly reflects the direct illumination range of non-directional luminaires within the channel.

[0030] Using the constructed direct illumination area map as a spatial reference, a comparative analysis is performed on each specific inspection target location extracted from the inspection target information obtained in step S2. Each inspection target location is checked to ensure it falls within the effective illumination range represented by the direct illumination area map. For inspection target locations outside the effective illumination range, they are identified as locations that cannot be fully illuminated directly by non-directional lighting fixtures, i.e., direct illumination blind spots. All identified direct illumination blind spot locations are compiled and specifically marked on a new map corresponding to the passage space, thus generating a map clearly indicating which key inspection points lack direct illumination. This map provides a clear input target and spatial basis for subsequent steps using directional lighting fixtures and reflected light for targeted supplementary lighting.

[0031] S4: Based on the lamp information and the reflective device sequence, the location and range of the reflection area are obtained through the lamp light reflection model; The position of the reflective device is obtained through the reflective device sequence. Based on the position of the directional lamp, the position of the non-directional lamp, the lamp parameters and the position of the reflective device, the position of the reflection area and the reflection range are solved by the lamp light reflection model. The position of the reflection area is taken as the position of the center point of the reflection area. The formula for the light reflection model of a lamp is: ; In the formula, For lighting fixtures Through the device In position Illuminance of reflected light, For lighting fixtures dimming ratio, For equipment Surface reflectivity, For lighting fixtures In the equipment Incident direction and equipment The angle between the normals, For equipment Arrive at the location The square of the distance, For equipment The effective reflective area, To pass through the device The reflected light at position The cosine of the angle between the normal of the surface receiving light and the incident direction; ; In the formula, For lighting fixtures Through the device In position The radius of the resulting reflection range, For equipment Arrive at the location distance, The equivalent divergence half-angle of the lamp The tangent value, This is the scaling factor from dimension to radius.

[0032] It should be further explained that the specific implementation process of step S4 is as follows: Based on the reflective device sequence constructed in step S1, the coordinates of each reflective device in the channel space are extracted, i.e., the reflective device position. Subsequently, the directional and non-directional luminaire positions contained in the luminaire information obtained from step S3 are used as spatial input parameters, and a preset luminaire light reflection model is called for calculation and analysis. The purpose is to solve for the spatial attributes of the lighting area formed by the reflection of luminaire light through the reflective devices, i.e., the reflection area position and reflection range.

[0033] The formula for the light reflection model of luminaires is used to calculate the indirect lighting effect caused by reflective devices within a channel. Its core consists of two parts: reflected illuminance calculation and reflection range estimation. The reflected illuminance part describes the illuminance value produced when light emitted by the luminaire reaches a spatial location after being reflected by the reflective device. This is the cosine of the angle between the reflected ray and the normal to the illuminated surface at the target point, and it is dimensionless. The reflection range part estimates the spatial influence of the reflected light spot. It is the reflective equipment to the evaluation location. The distance; It is the tangent of the equivalent divergence half-angle of the lamp, dimensionless, and characterizes the natural diffusion of the light beam; This is an empirical proportionality coefficient, dimensionless, that converts the size of reflective devices into radius addition. The model as a whole quantifies the secondary distribution of light after reflection, serving as a crucial computational basis for distinguishing between beneficial supplementary lighting and harmful interference. In practical applications, the reflected illuminance is calculated for each pair of luminaires and reflective devices, and for a large number of points within the meshed channel. For a specific reflective device, all spatial points that contribute significantly to illuminance within the channel due to their reflection are identified. The set of these points constitutes a roughly continuous reflected illumination area. The spatial geometric center of this area is calculated, and the coordinates of this center point are defined as the location of the reflection area corresponding to this reflection. Simultaneously, based on the radius of the reflection range calculated by the model, combined with the location of the reflection area, a boundary describing the spatial extent of the reflected light's influence is determined, i.e., the reflection range. This range is typically characterized as a spatial domain centered on the location of the reflection area and with the radius of the reflection range as the scale of influence.

[0034] The output is a data set of locations of a series of reflection areas and their corresponding reflection ranges generated by the interaction of all luminaires and all reflective devices. This data precisely quantifies the spatial distribution and intensity of the secondary lighting effect produced by indirect reflection within the channel, and is a key input for distinguishing the gain and interference effects of reflected light in subsequent steps.

[0035] S5: Obtain the inspection plan data of the inspection robot, and construct a reflection gain interference conversion sequence based on the inspection plan data, the location of the reflection area, and the reflection range; The inspection plan data of the inspection robot is acquired, and natural language processing technology is used to process the inspection plan data to obtain planned inspection target information and planned inspection location information. Based on the planned inspection target information, the planned inspection target location is obtained, and based on the planned inspection location information, the planned detection sensor location is obtained. The reflection range is compared with the planned detection sensor location and the planned inspection target location. If the planned detection sensor location is within the reflection range, the gain-interference conversion coefficient is calculated, and the gain or interference state is marked based on the gain-interference conversion coefficient. If the planned detection sensor location is not within the reflection range but the planned inspection target location is within the reflection range, it is marked as a gain state. Based on the marking results, a reflection-gain-interference conversion sequence is generated.

[0036] The formula for the gain-interference conversion factor is: ; In the formula, For gain-to-interference conversion coefficient, The relative gain due to reflection. Interference intensity; ; In the formula, For reflected illuminance, To detect the safe illuminance threshold of the sensor; ; In the formula, For direct illuminance, It is a very small positive number.

[0037] It should be further explained that the specific implementation process of step S5 is as follows: Figure 2 As shown, future inspection plan data is obtained from the inspection robot's scheduling management platform or task planning system. This data is typically described in natural language as task instructions or plan documents, which contain the inspection tasks that the robot will perform.

[0038] Subsequently, natural language processing (NLP) technology was used to automatically parse and structure the acquired inspection plan data. This process aimed to extract two key types of planning information from the plan text: planned inspection target information and planned inspection location information. The planned inspection target information clearly defines the list of equipment objects that need to be inspected in the upcoming inspection task and their related requirements; the planned inspection location information describes the expected movement trajectory and working points of the robot and its onboard sensors in the planned task.

[0039] Based on the analyzed planned inspection target information, the specific coordinates of each planned inspection target in the three-dimensional space of the channel are further extracted, i.e., the planned inspection target position. Simultaneously, based on the analyzed planned inspection position information, the precise spatial coordinates of the various optical detection sensors carried by the inspection robot when performing detection actions are extracted, i.e., the planned detection sensor positions.

[0040] The reflection range corresponding to each reflection area calculated in step S4 is spatially compared and analyzed with the planned detection sensor location and the planned inspection target location. This comparison process determines the nature of the impact of reflected light on the upcoming inspection operation. Each reflection range is assessed individually. The first case is if a planned detection sensor location falls within the current reflection range, meaning the reflected light can directly illuminate the sensor. In this case, the gain-interference conversion coefficient is calculated to quantify the degree of benefit or harm of the reflection's impact. The formula for calculating the gain-interference conversion coefficient involves two core variables: the relative gain brought by the reflection and the interference intensity. The gain-interference conversion coefficient equals the relative gain divided by one and the sum of the interference intensity. The reflected illuminance is calculated using the lamp light reflection model in step S4. The safe illuminance threshold is a pre-set maximum safe illuminance value that ensures the sensor is not saturated or damaged by excessive light interference. After calculating the gain-interference conversion coefficient, it is marked as either a gain state or an interference state based on its numerical range. For example, when the coefficient is greater than a certain threshold, it is marked as a gain state, indicating that the benefits of reflected light outweigh the drawbacks; conversely, it is marked as an interference state.

[0041] The second scenario involves a planned inspection target location falling within the planned reflection range of a sensor that is not currently within that range. This means the reflected light can illuminate the equipment to be inspected without directly interfering with the sensor. In this case, the reflection effect is marked as a gain state. After comparing and marking all reflection ranges and planned locations, all marking results are organized and sorted according to the reflection area, the corresponding luminaire, and the combination of reflective devices, generating a structured sequence—the reflection gain-interference conversion sequence. This sequence clearly records whether each potential reflected light effect, under a specific future inspection plan, serves as a beneficial supplementary lighting gain or an interference that needs to be suppressed, providing a direct decision-making basis for the final lighting adjustment and control.

[0042] S6: Based on the lamp information, the target position map of the direct blind spot, and the reflective device sequence, solve the directional lamp adjustment sequence; based on the lamp information and the reflection gain interference conversion sequence, solve the non-directional lamp adjustment sequence; and perform lighting adjustment control based on the non-directional lamp adjustment sequence and the directional lamp adjustment sequence.

[0043] Based on the planned inspection target location, the required illuminance for the planned inspection blind spot target location is obtained. The directional lighting fixture adjustment sequence is optimized by combining the directional lighting fixture locations, reflector locations, and the direct blind spot target location map. A reflection gain interference conversion sequence is generated based on the planned inspection target location, and the non-directional lighting fixture adjustment sequence is optimized by combining the non-directional lighting fixture locations. Through the lighting fixture control unit, precise adjustment commands, including azimuth adjustment, pitch adjustment, and dimming ratio adjustment, are sent to each directional lighting fixture according to the directional lighting fixture adjustment sequence. A dimming ratio adjustment command is sent to each non-directional lighting fixture according to the non-directional lighting fixture adjustment sequence.

[0044] Formula for the objective function of directional lighting optimization: ; In the formula, To optimize the objective function value for directional lighting fixtures, Blind spot target location The required illuminance, For the number of directional lights, For directional lighting fixtures Target location in the blind spot Illuminance of reflected light, For directional lighting fixtures The square of the azimuth adjustment amount, For directional lighting fixtures The square of the pitch angle adjustment amount, For directional lighting fixtures The square of the dimming ratio adjustment amount. and These are the weighting coefficients; Formula for the objective function of non-directional lighting fixture optimization: ; In the formula, Optimize the objective function value for non-directional luminaires. For the inspection target location The required illuminance, For the inspection target location The sum of direct illuminance and reflected illuminance, Non-directional lighting fixtures The square of the dimming ratio adjustment amount. At the inspection target location Gain-interference conversion coefficient at the location, , and These are the weighting coefficients.

[0045] It should be further explained that the specific implementation process of step S6 is as follows: First, based on the planned inspection target locations determined in steps S2 and S5, select the locations within the range of the direct blind spot target location map generated in step S3. These points are defined as planned inspection blind spot target locations. Obtain the preset lighting level corresponding to these specific locations from the inspection target information to meet effective detection, i.e., the required illuminance of the planned inspection blind spot target locations.

[0046] Subsequently, an optimization process is performed for the directional luminaires to generate an adjustment sequence. The core inputs to this process include: the locations of all directional luminaires extracted from the luminaire information, the locations of all reflectors extracted from the reflector sequence, and a map identifying the target locations of blind spots without direct illumination. The optimization goal is to calculate an optimal set of adjustment parameters for each directional luminaire, ensuring that the emitted light, after reflection by the reflectors, accurately reaches the planned inspection blind spot target location, and that the reflected illuminance at that location is close to its required illuminance, while minimizing the luminaire's own adjustment range to reduce control costs.

[0047] Solve the objective function for directional lighting optimization. The formula for the objective function for directional lighting optimization consists of two main weighted parts. The value of the objective function for directional lighting optimization is unitless, and all parameters are dimensionless before calculation; the smaller the value, the lower the adjustment cost of the directional lighting. The first part is the illuminance deviation term, which reflects the difference between the required illuminance and the total reflected illuminance produced by all directional lighting at a certain blind spot target location. The second part is the adjustment cost term, which is the sum of the squares of the azimuth adjustment, the squares of the pitch adjustment, and the squares of the dimming ratio adjustment of the directional lighting. The process of determining the weight coefficients involves inviting experts in the lighting and inspection fields to compare each criterion in the optimization objective pairwise, quantifying their relative importance using the 1-9 scale, and constructing a judgment matrix. For example, for the illuminance satisfaction and adjustment cost criteria of directional lighting optimization, the largest eigenvalue of the matrix and its corresponding normalized eigenvector are then calculated. Each component of this eigenvector is the initial weight coefficient of each criterion. Consistency checks are required during the calculation process to ensure the logical consistency of the expert judgments, requiring a consistency ratio of less than 0.1. This process yields a set of values, such as 0.75 (illuminance deviation weight) and 0.25 (adjustment cost weight), for directional lighting optimization. By minimizing the objective function value through an optimization algorithm, the optimal azimuth adjustment, pitch adjustment, and dimming ratio adjustment for each directional lighting fixture are determined. These solutions together constitute the directional lighting fixture adjustment sequence.

[0048] Simultaneously, optimization solutions for non-directional luminaires are performed in parallel to obtain a non-directional luminaire adjustment sequence. This optimization is based on the reflection gain interference conversion sequence generated in step S5 and the non-directional luminaire positions in the luminaire information. The goal is to adjust the dimming ratio of all non-directional luminaires to achieve sufficient and high-quality overall illumination at key planned inspection locations, while controlling the adjustment range of the luminaires.

[0049] Solve for an objective function to optimize non-directional lighting fixtures. The value of the objective function is a weighted sum of three parts. The value of the objective function is unitless, and all parameters are dimensionless before calculation. The smaller the value, the lower the adjustment cost of the non-directional lighting fixtures. The first part is the illuminance deficiency penalty term. For a certain inspection target location, when the sum of direct illuminance and reflected illuminance at that location is lower than its required illuminance, the difference is calculated and the maximum value is taken; if the sum of direct illuminance and reflected illuminance already meets the requirements, this term is zero. The sum of direct illuminance and reflected illuminance is obtained by superimposing the direct illuminance received from all non-directional lighting fixtures at that location with all beneficial reflected illuminance. The second part is the control cost term, which is the sum of the squares of the dimming ratio adjustments of all non-directional lighting fixtures. The third part is the gain utilization term, which is related to the gain-interference conversion coefficient corresponding to the inspection target location. The gain-interference conversion coefficient used in this calculation is normalized and its value is between 0 and 1. A larger coefficient indicates a better reflected light gain effect. Therefore, a negative sign is used before this term or it is treated as a subtraction term in the function to encourage gain utilization. The weighting coefficient setting process also first uses the analytic hierarchy process (AHP) to determine the initial values. Experts are invited to conduct pairwise comparisons of the three criteria: "meeting the inspection target illuminance," "reducing the cost of lamp adjustment," and "utilizing gain and suppressing interference." A judgment matrix is ​​constructed, and the normalized eigenvector is calculated. Its components are the recommended initial values ​​for the weighting coefficients. Typical initial values ​​are as follows: It ranges from 0.5 to 0.7. It ranges from 0.1 to 0.2. The value is 0.2 to 0.3. Subsequently, during operation, online fine-tuning is performed based on feedback from three performance indicators: the target illuminance compliance rate, the total dimming amplitude change of non-directional lamps, and the sensor interference rate. The objective function value is minimized through an optimization algorithm, thereby solving for the optimal dimming ratio adjustment amount for each non-directional lamp. These solutions together constitute the non-directional lamp adjustment sequence.

[0050] In the final execution stage of lighting regulation control, the results of collaborative optimization are transformed into specific executable physical commands. The core of this process lies in integrating, synchronizing, and distributing the calculated directional luminaire adjustment sequences with the non-directional luminaire adjustment sequences, driving the luminaire group to reconstruct the lighting environment. The specific implementation process is as follows: The control center first merges the two adjustment sequences to generate a complete lighting control instruction set. This instruction set is a structured data packet that not only contains the final target parameters for each luminaire, but also specifies the order of instruction execution and coordination timing to ensure that the supplementary lighting and global lighting effects are presented smoothly and synchronously.

[0051] Command issuance and execution are accomplished through the control unit of the lighting system (such as DALI, 0-10V, or IP-based smart lighting protocol gateway). For each directional luminaire, the control unit sends it a triplet command, including: Target azimuth: the absolute angle of horizontal rotation of the luminaire; Target pitch: the absolute angle of vertical tilt of the luminaire; Target dimming ratio: the percentage of output power of the luminaire's light source. For each non-directional luminaire, a command containing the target dimming ratio is sent. Example: Adjustment sequence for directional luminaires: Light fixture D_Light_01: Azimuth angle adjusted to 35.2°, pitch angle adjusted to -12.5°, dimming ratio increased to 85%. Its goal is to use the high-reflectivity cable grounding box located at (X1,Y1,Z1) to provide supplemental lighting for the target in the dead angle, Target_A (a cable connector located on the back of the bracket).

[0052] Non-directional lighting adjustment sequence: Light ND_Light_07: Dimming ratio reduced from 60% to 45%. The direct light from this light fixture is one of the sources of interference causing overexposure of the inspection robot's gimbal camera at path point Path_Point_3.

[0053] Once all luminaires report that they have reached the target state set by the command, the lighting field of the high-voltage cable channel is dynamically reconstructed. At this point, the secondary light source formed by the reflection of the reflective equipment accurately covers the predetermined blind spots of the inspection. At the same time, the uniform background lighting provided by the non-directional luminaires meets the global illuminance requirements of the inspection path and effectively suppresses the risk of overexposure at the sensor location, thus achieving coordinated control of precise local supplementary lighting and global optimization and interference suppression.

[0054] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A data analysis-based intelligent control method for lighting in high-voltage cable tunnels, characterized in that, Includes the following steps: S1: Obtain equipment information for the high-voltage cable channel, and filter to obtain a sequence of reflective devices based on the equipment information; S2: Process historical operation and maintenance data of high-voltage cable channels to obtain inspection target information; S3: Obtain the lighting information of the high-voltage cable channel, construct a direct lighting area map based on the lighting information, and construct a direct lighting blind spot target location map based on the direct lighting area map and the inspection target information; S4: Based on the lamp information and the reflective device sequence, the location and range of the reflection area are obtained through the lamp light reflection model; S5: Obtain the inspection plan data of the inspection robot, and construct a reflection gain interference conversion sequence based on the inspection plan data, the location of the reflection area, and the reflection range; S6: Based on the lamp information, the target position map of the direct blind spot, and the reflective device sequence, solve the directional lamp adjustment sequence; based on the lamp information and the reflection gain interference conversion sequence, solve the non-directional lamp adjustment sequence; and perform lighting adjustment control based on the non-directional lamp adjustment sequence and the directional lamp adjustment sequence.

2. The intelligent control method for high-voltage cable channel lighting based on data analysis according to claim 1, characterized in that, The process of obtaining equipment information for high-voltage cable channels and filtering a sequence of reflective devices based on the equipment information includes: the equipment information including equipment identifier, equipment type, equipment location, effective reflective area, and surface reflectivity; dividing the devices into reflective and non-reflective devices based on the surface reflectivity; and constructing a sequence of reflective devices by obtaining the equipment identifier and surface reflectivity of the reflective devices.

3. The intelligent control method for high-voltage cable channel lighting based on data analysis according to claim 1, characterized in that, The process of processing historical operation and maintenance data of high-voltage cable channels to obtain inspection target information includes: acquiring historical operation and maintenance data of high-voltage cable channels, performing natural language processing on the historical operation and maintenance data to obtain inspection target information, wherein the inspection target information includes the required illumination of the inspection target and the location of the inspection target.

4. The intelligent control method for high-voltage cable channel lighting based on data analysis according to claim 1, characterized in that, The process of acquiring lighting information of the high-voltage cable channel, constructing a direct illumination area map based on the lighting information, and constructing a direct illumination blind spot target location map based on the direct illumination area map and the inspection target information includes: the lighting information includes directional lighting positions, non-directional lighting positions, and lighting parameters; the high-voltage cable channel area is gridded; the illuminance of each grid is solved using a direct illumination model based on the non-directional lighting positions and lighting parameters to construct the direct illumination area map; and the direct illumination blind spot target location is determined based on the direct illumination area map and the inspection target location to generate the direct illumination blind spot target location map.

5. The intelligent control method for high-voltage cable channel lighting based on data analysis according to claim 4, characterized in that, The step of solving the direct illumination area using a direct illuminance model based on the non-directional luminaire position and luminaire parameters includes: the direct illuminance model formula is: ； In the formula, Non-directional lighting fixtures In position direct illuminance, Non-directional lighting fixtures In the main axis and position of the lamp included angle Light intensity distribution, Non-directional lighting fixtures Arrive at the location The square of the distance, Non-directional lighting fixtures In position The cosine of the angle between the normal of the illuminated surface and the incident direction. Non-directional lighting fixtures The dimming ratio.

6. The intelligent control method for high-voltage cable channel lighting based on data analysis according to claim 4, characterized in that, The step of obtaining the reflection area position and reflection range based on the lamp information and the reflector sequence through the lamp light reflection model includes: obtaining the reflector position through the reflector sequence, and solving the reflection area position and reflection range based on the directional lamp position, non-directional lamp position, lamp parameters and reflector position through the lamp light reflection model, wherein the reflection area position is taken as the center point position of the reflection area; The formula for the light reflection model of a lamp is: ； In the formula, For lighting fixtures Through the device In position Illuminance of reflected light, For lighting fixtures dimming ratio, For equipment Surface reflectivity, For lighting fixtures In the equipment Incident direction and equipment The angle between the normals, For equipment Arrive at the location The square of the distance, For equipment The effective reflective area, To pass through the device The reflected light at position The cosine of the angle between the normal of the surface receiving light and the incident direction; ； In the formula, For lighting fixtures Through the device In position The radius of the resulting reflection range, For equipment Arrive at the location distance, The equivalent divergence half-angle of the lamp The tangent value, This is the scaling factor from dimension to radius.

7. The intelligent control method for high-voltage cable channel lighting based on data analysis according to claim 1, characterized in that, The process of acquiring the inspection plan data of the inspection robot and constructing a reflection gain-interference conversion sequence based on the inspection plan data, the reflection area location, and the reflection range includes: acquiring the inspection plan data of the inspection robot; processing the inspection plan data using natural language processing technology to obtain planned inspection target information and planned inspection location information; obtaining the planned inspection target location based on the planned inspection target information; obtaining the planned detection sensor location based on the planned inspection location information; comparing the reflection range with the planned detection sensor location and the planned inspection target location; if the planned detection sensor location is within the reflection range, calculating the gain-interference conversion coefficient; marking the gain or interference state based on the gain-interference conversion coefficient; if the planned detection sensor location is not within the reflection range but the planned inspection target location is within the reflection range, marking it as a gain state; and generating a reflection gain-interference conversion sequence based on the marking results.

8. A data analysis-based intelligent control method for lighting in high-voltage cable tunnels according to claim 7, characterized in that, If the planned sensor is located within the reflection range, then a gain-interference conversion coefficient is calculated, and the gain or interference state is marked based on the gain-interference conversion coefficient, including: The formula for the gain-interference conversion factor is: ； In the formula, For gain-to-interference conversion coefficient, The gain intensity brought about by reflection Interference intensity; ； In the formula, For reflected illuminance, To detect the safe illuminance threshold of the sensor; ； In the formula, For direct illuminance, It is a very small positive number.

9. A data analysis-based intelligent control method for lighting in high-voltage cable tunnels according to claim 1, characterized in that, The process of solving the directional lighting adjustment sequence based on the lighting fixture information, the target location map of the direct blind spot, and the reflector sequence, and solving the non-directional lighting adjustment sequence based on the lighting fixture information and the reflection gain interference conversion sequence, and then combining the directional lighting adjustment sequence for lighting adjustment control, includes: obtaining the required illuminance of the target location of the planned inspection blind spot based on the planned inspection target location, and optimizing the directional lighting adjustment sequence by combining the directional lighting fixture location, the reflector location, and the target location map of the direct blind spot; generating the reflection gain interference conversion sequence based on the planned inspection target location, and optimizing the non-directional lighting adjustment sequence by combining the non-directional lighting fixture location; and sending precise adjustment commands, including azimuth adjustment, pitch adjustment, and dimming ratio adjustment, to each directional lighting fixture through the lighting fixture control unit according to the directional lighting adjustment sequence, and sending dimming ratio adjustment commands to each non-directional lighting fixture according to the non-directional lighting adjustment sequence.

10. The intelligent control method for high-voltage cable channel lighting based on data analysis according to claim 1, characterized in that, The method combines the location of the directional lighting fixtures, the location of the reflective devices, and the location map of the target in the direct blind spot to optimize and solve the directional lighting fixture adjustment sequence; Based on the planned inspection target location, a reflection gain interference conversion sequence is generated. Combined with the non-directional luminaire location optimization, a non-directional luminaire adjustment sequence is solved, including: Formula for the objective function of directional lighting optimization: ； In the formula, To optimize the objective function value for directional lighting fixtures, Blind spot target location The required illuminance, For the number of directional lights, For directional lighting fixtures Target location in the blind spot Illuminance of reflected light, For directional lighting fixtures The square of the azimuth adjustment amount, For directional lighting fixtures The square of the pitch angle adjustment amount, For directional lighting fixtures The square of the dimming ratio adjustment amount. and These are the weighting coefficients; Formula for the objective function of non-directional lighting fixture optimization: ； In the formula, Optimize the objective function value for non-directional luminaires. For the inspection target location The required illuminance, For the inspection target location The sum of direct illuminance and reflected illuminance, Non-directional lighting fixtures The square of the dimming ratio adjustment amount. At the inspection target location Gain-interference conversion coefficient at the location, , and These are the weighting coefficients.

Images