Distribution box room environment parameter integrated measurement method

Abstract

The invention discloses a distribution box room environment parameter integrated measurement method, and particularly relates to the field of multi-parameter measurement, and the method comprises the steps: firstly constructing a machine room three-dimensional model and a sensor correlation degree matrix, then building a heat balance, humidity diffusion and airflow motion model, and calibrating parameters; deploying a measurement system and calibrating through dual synchronization and cross check; establishing a sensor confidence evaluation system and an adaptive threshold based on 72-hour reference data; fusing calibration data by adopting a three-level correlation calibration mechanism; four types of measurement modes and conversion logics are designed, and measurement resources are dynamically allocated through environmental risk assessment; multi-dimensional state parameters are extracted, a comprehensive evaluation value is calculated through normalization, dynamic weighting and combinatorial algorithms, and five-level early warning response is achieved; according to the method, multiple models and an intelligent algorithm are integrated, the parameter measurement precision and the environment risk identification efficiency are improved, the response delay is reduced, the fault diagnosis accuracy is improved, and reliable technical support is provided for safe operation and maintenance of the distribution box room.

Description

Technical Field

[0001] This invention relates to the field of multi-parameter measurement technology, and more specifically, to an integrated method for measuring environmental parameters of a power distribution box room. Background Technology

[0002] As a critical link in power transmission and distribution systems, the environmental parameters of distribution rooms directly affect the operational stability and lifespan of power equipment. Environmental monitoring is a core means of ensuring power supply reliability. Existing technologies include multi-parameter monitoring schemes for distribution rooms, which collect environmental data from key areas by deploying various sensing devices such as temperature, humidity, and vibration sensors. Data transmission primarily utilizes wired or wireless communication technologies to transmit collected data to terminal platforms. Some schemes introduce basic data processing and early warning mechanisms, judging environmental conditions based on preset thresholds to provide basic references for operation and maintenance. Simultaneously, existing technologies have attempted to combine simple mathematical models or statistical methods to conduct preliminary analysis of monitoring data to meet basic environmental monitoring and safety assurance needs.

[0003] However, it still has some drawbacks in practical use, such as:

[0004] 1. Sensor placement relies heavily on human experience and has not established a spatial correlation system based on heat conduction, airflow path, and physical distance. The data from each sensor are isolated and lack coordination, making it impossible to use correlation to achieve data cross-validation and anomaly diagnosis, resulting in insufficient data reliability.

[0005] 2. Most models use a single mathematical model or a fixed parameter model, without considering the coupling relationship of environmental parameters. Furthermore, the model parameters lack a dynamic correction mechanism, making it difficult to accurately adapt to the complex and ever-changing heat flow and airflow environment of the power distribution room, resulting in large measurement errors.

[0006] 3. Time synchronization often relies on a single software protocol, with synchronization accuracy only reaching the millisecond level or above. Furthermore, it lacks a graded calibration mechanism and can only remove outliers through simple thresholds, which cannot effectively correct sensor drift and system deviations, thus affecting data accuracy.

[0007] 4. The sampling rate is fixed and measurement resources are not dynamically allocated in conjunction with environmental risks. High-risk areas may have insufficient monitoring frequency, while low-risk areas may experience energy waste. Furthermore, the assessment focuses only on a single parameter and lacks multi-dimensional comprehensive judgment, resulting in low accuracy of early warning. Summary of the Invention

[0008] In order to overcome the above-mentioned defects of the prior art, the present invention provides an integrated measurement method for environmental parameters of power distribution box rooms, which solves the problems mentioned in the background art through the following scheme.

[0009] To achieve the above objectives, the present invention provides the following technical solution: an integrated method for measuring environmental parameters of a power distribution box room, comprising:

[0010] S1: Overall design of the measurement system: Perform three-dimensional topology planning for multi-parameter measurement points, select and adapt sensors and construct sensor correlation matrix based on the correspondence between measurement points and parameters;

[0011] S2: Construction and parameter calibration of multivariate correlation measurement model: Establish thermal equilibrium model, humidity diffusion model and airflow motion model, build a calibration system through a controllable environment chamber, perform experiments to collect data by parameter division, fit model parameters and establish a correction function library;

[0012] S3: Measurement System Deployment and Synchronization Calibration: Install sensors and measurement units, configure master-slave communication links, achieve full system time synchronization through hardware triggering and software protocols, build a cross-validation network and perform initial calibration;

[0013] S4: Measurement benchmark establishment and self-verification system initialization: Execute 72 hours of continuous data acquisition under different working conditions, process the data to obtain parameter statistical characteristics, correlation matrix and normal fluctuation range, verify the model accuracy, establish a sensor confidence evaluation system and set adaptive diagnostic thresholds;

[0014] S5: Multi-sensor data fusion and dynamic calibration: Preprocessing and time alignment of the acquired data, Kalman filtering for noise reduction, and evidence theory fusion of data based on sensor confidence. Dynamic calibration of data is achieved through a three-level correlation calibration mechanism and the calibration results are fed back.

[0015] S6: Measurement and Control: Defines four types of measurement modes: normal, early warning, fault, and maintenance. Designs mode conversion logic based on multi-parameter joint criteria and realizes data acquisition resource allocation through environmental risk assessment.

[0016] S7: Comprehensive Measurement Status Assessment and Early Warning: Extract multi-dimensional status parameters such as heat, humidity, airflow and physical environment, use normalization processing, dynamic weight allocation and combination algorithm to calculate comprehensive assessment value, analyze status trend and execute early warning response according to level.

[0017] The technical effects and advantages of this invention are as follows:

[0018] 1. Construct a three-dimensional sensor correlation matrix that integrates heat conduction, airflow path, and physical distance, clarify the master-slave relationship and data collaboration logic, break the isolation of sensor data, provide a foundation for cross-validation and fusion analysis, and significantly improve data correlation and reliability.

[0019] 2. Establish a multi-model fusion system for thermal balance, humidity diffusion, and airflow motion. Through the calibration and dynamic correction mechanism of the controllable environment chamber system, accurately characterize the parameter coupling relationship, control the relative error between the model calculation value and the measured value, and effectively adapt to the parameter change law of complex environment.

[0020] 3. Employing both hardware triggering and PTPv2 protocol for dual time synchronization, a three-level correlation calibration mechanism is designed. By combining theoretical values ​​from the model with data from nearby sensors to dynamically correct deviations, system errors are significantly reduced, ensuring industrial-grade measurement accuracy.

[0021] 4. Through environmental risk assessment, measurement resources are dynamically allocated. The sampling rate is automatically increased and backup sensors are activated in high-risk areas, while energy consumption is optimized in low-risk areas. A five-level accurate early warning is achieved through multi-dimensional comprehensive assessment, balancing monitoring effectiveness and energy consumption. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the overall structure of the present invention.

[0023] Figure 2 This is a schematic diagram of the sensor correlation matrix construction process of the present invention.

[0024] Figure 3 This is a schematic diagram of the data fusion and calibration process of the present invention.

[0025] Figure 4 This is a schematic diagram of the status assessment and early warning process of the present invention. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] refer to Figures 1-4 The method for integrated measurement of environmental parameters in a power distribution box room, as shown, includes:

[0028] S1: Overall Design of the Measurement System: Perform three-dimensional topology planning for multi-parameter measurement points, select and adapt sensors based on the correspondence between measurement points and parameters, and construct a sensor correlation matrix; the specific steps are as follows:

[0029] S101: 3D topology planning for multi-parameter measurement points:

[0030] Using BIM technology combined with laser scanning, a 1:1 three-dimensional spatial model of the power distribution room was constructed. The model includes equipment layout, ventilation system and building structure. The three-dimensional spatial coordinate system was determined. With the center point of the equipment room floor as the origin, a right-hand rectangular coordinate system with X-axis horizontal, Y-axis vertical and Z-axis height was established.

[0031] Based on the simulation analysis of the thermal flow field and airflow field of the power distribution box room, the key measurement locations were determined: the surface of the main heating element, the ventilation area of ​​the air inlet and outlet, the bottom space 30cm above the ground, the top space 30cm below the ceiling, the cable inlet and outlet, and the environmental reference point in the middle of the computer room.

[0032] Based on the environmental characteristics of the measurement location, the type of measurement parameters for each point is determined, and a correspondence table between measurement points and parameters is formed.

[0033] S102: Sensor Selection and Correlation Matrix Construction:

[0034] Based on the correspondence table of measurement points and parameters, select and match the following sensors: PT1000 platinum resistance temperature sensor for the surface of the heating element, SHT35 humidity sensor and FS4000 wind speed sensor for the ventilation area, IEPE vibration sensor for the cable inlet and outlet, and integrated temperature, humidity, air pressure and PM2.5 sensor for the middle of the computer room.

[0035] Define the dimensions of correlation between sensors:

[0036] thermal conductivity correlation : Where k is the thermal conductivity of the material, and A is the heat exchange area at the location of the two sensors. d represents the historical average temperature difference, and d is the physical distance between the two sensors.

[0037] airflow path correlation The probability of airflow between two sensors was obtained through CFD simulation: 0.8-1.0 for direct coverage, 0.3-0.7 for indirect coverage, and 0-0.2 for no coverage.

[0038] Physical distance correlation :according to Normalization, in which The distance between the two sensors in three-dimensional space is Euclidean distance, in meters.

[0039] Calculate the overall correlation: Construct an n×n sensor correlation matrix, where n is the total number of sensors, and clarify the master-slave relationship, with the sensors integrated in the middle of the computer room as the master unit and the rest as slave units;

[0040] It should be further explained that, in this comprehensive correlation function, since local overheating of the heating element is the main risk, the heat conduction correlation has the highest weight, at 0.4. Airflow is the main carrier for the diffusion of parameters such as temperature, humidity, and particulate matter in the computer room, directly affecting the parameter correlation of non-heating areas. Physical distance is the basic reference for parameter correlation. The closer the distance, the closer the parameters are theoretically. However, in the complex environment of the computer room, its influence is easily offset by other factors. Therefore, the airflow path correlation and physical distance correlation have the second highest weights, both at 0.3.

[0041] S2: Construction and Parameter Calibration of Multivariable Correlation Measurement Model: A thermal equilibrium model, a humidity diffusion model, and an airflow motion model are established. A calibration system is built using a controllable environment chamber. Experiments are performed to collect data for each parameter, model parameters are fitted, and a correction function library is established. The specific steps are as follows:

[0042] S201: Construct the core mathematical models, including: thermal equilibrium model, humidity diffusion model, and airflow motion model, as detailed below:

[0043] The thermal equilibrium model describes the temperature change of the heating element, and its specific mathematical function is as follows:

[0044] ;

[0045] in, This is the heat accumulation term, representing the heat that is not dissipated from the component per unit time. It is the direct cause of the component's temperature rise, and its specific mathematical function is: Where c is the specific heat capacity of the component material, and m is the mass of the component. The change in temperature For time intervals;

[0046] This is the heat generation term, representing the heat generated by the component due to electrical energy loss per unit time. It is the sole source of the component's own heat, and its specific mathematical function is: Where P is the rated power of the component and η is the component efficiency;

[0047] This is the convective heat dissipation term, representing the heat lost from components to the environment per unit time through fluid flow. It is the main way components in the distribution box are cooled, and its specific mathematical function is: Where h is the convective heat transfer coefficient and A is the heat dissipation area of ​​the element. The surface temperature of the component. The ambient temperature;

[0048] This is the radiation heat dissipation term, representing the heat lost by the component to the surrounding environment through thermal radiation per unit time. Its specific mathematical function is: ,in, The emissivity of the component surface, It is the Stefan-Boltzmann constant. The fourth power of the thermodynamic temperature;

[0049] A humidity diffusion model describes the humidity distribution in a computer room, and its specific mathematical function is as follows:

[0050] Where D is the humidity diffusion coefficient. , For the humidity Laplace operator, The three second-order partial derivatives correspond to the rates of change of humidity gradients in the X, Y, and Z directions, respectively. For humidity source items, , where represents the rate of change in humidity per unit volume of the computer room air caused by external factors, and V is the ventilation volume of the computer room ventilation system. For air intake humidity, For air humidity, This refers to the total volume of the computer room;

[0051] An airflow motion model describes the relationship between the airflow velocity at the ventilation opening and influencing factors. Its specific mathematical function is as follows:

[0052] Where v is the wind speed, n is the wind turbine speed, and k, a, and b are model parameters.

[0053] It should be further explained that the calibration of the above model parameters specifically includes:

[0054] Set up a calibration experimental system, including a controllable environment chamber (temperature -10-50℃, humidity 20%-95%RH, wind speed 0-5m / s), a simulated load consistent with the actual computer room (adjustable power circuit breaker, transformer), and high-precision data acquisition equipment (1MHz oscilloscope, ±0.05℃ data logger).

[0055] Perform calibration experiments using different parameters:

[0056] Temperature parameter calibration: Adjust the ambient chamber temperature (-10-50℃) in 5℃ increments, collect the surface temperature of each heating element and the ambient temperature, record the element power and heat dissipation area, and fit the h and ε parameters in the thermal balance model using the least squares method.

[0057] Humidity parameter calibration: The humidity of the environmental chamber was adjusted in 10% RH increments (20%-95% RH), humidity values ​​were collected at different spatial locations, the humidity gradient was calculated, and the D value in the humidity diffusion model was fitted. parameter;

[0058] Wind speed parameter calibration: Adjust the fan speed in 100r / min increments (500-2000r / min), set different temperature differences inside and outside the machine room (5-20℃), collect the wind speed at the ventilation outlet, and fit the k, a, and b parameters in the airflow motion model;

[0059] Establish a parameter correction function library: record the model parameter values ​​under different environmental conditions, form a fitting function for parameters as they change with the environment, and embed it into the subsequent data processing module;

[0060] S3: Measurement System Deployment and Synchronization Calibration: Install sensors and measurement units, configure master-slave communication links, achieve system-wide time synchronization through hardware triggering and software protocols, build a cross-validation network, and perform initial calibration; the specific steps are as follows:

[0061] S301: Hardware System Deployment:

[0062] Sensor Installation: Temperature and Humidity Sensors: Fixed to a metal bracket with insulated screws. The bracket is insulated from the equipment housing. The sensor on the surface of the heating element should be in close contact with a non-energized area, and at least 5cm away from the wiring terminals.

[0063] Wind speed sensor: fixed to the vent via a universal joint to ensure that the angle between the probe axis and the airflow direction is less than 5°;

[0064] Vibration sensor: It is fixed to the surface of the cable inlet and outlet equipment with a magnetic base, with a fit of ≥90%, avoiding welds and bolts;

[0065] Measurement unit deployment:

[0066] The integrated sensor of the main measurement unit is installed in the power distribution box in the middle of the computer room, 1.2-1.5m above the ground, with the heat dissipation holes facing the ventilation direction;

[0067] The measurement unit is integrated with the corresponding sensor. The cable is shielded (length ≤ 5m), the shielding layer is grounded at one end, and the distance between it and the high-voltage cable is ≥ 0.5m to avoid parallel laying.

[0068] Communication link configuration: LoRaWAN communication is used between the master and slave units, and gigabit Ethernet or industrial-grade 4G is used between the master unit and the edge gateway;

[0069] S302: System Synchronization and Cross-calibration:

[0070] Time synchronization calibration:

[0071] Hardware-triggered synchronization: The master unit outputs a 5V optocoupler-isolated trigger signal to trigger all slave units to synchronously acquire data, with a trigger delay of less than 0.5ms;

[0072] Software protocol synchronization: The PTPv2 precise time protocol is adopted. The main unit synchronizes with the NTP server with ±1ms accuracy every hour to ensure that the time deviation of the entire system is less than 1ms.

[0073] Cross-validation network construction:

[0074] Two sensors of the same model are deployed at each measurement point to establish redundant channels;

[0075] Set the verification rules: When the deviation between the main and backup sensor data is ≤0.5℃ or 2%RH, the main sensor data is used; when the deviation is 0.5-1℃ or 2%-5%RH, the average value is taken; when the deviation is greater than 1℃ or greater than 5%RH, a fault warning is triggered and the backup sensor is activated.

[0076] System initial calibration: Collect the no-load output values ​​of each sensor, compare them with standard instruments, calculate correction coefficients, compensate for the sensor output values, and the error after correction is less than 1%;

[0077] S4: Measurement Benchmark Establishment and Self-Verification System Initialization: Perform 72 hours of continuous data acquisition under different operating conditions, process the data to obtain parameter statistical characteristics, correlation matrix, and normal fluctuation range, verify model accuracy, establish a sensor confidence evaluation system, and set adaptive diagnostic thresholds; the specific steps are as follows:

[0078] S401: Continuous Reference Data Acquisition and Processing

[0079] Setting up a data acquisition scheme based on different operating conditions: 72 hours of baseline data were collected.

[0080] No-load operation: Equipment power 0%, natural ventilation, data collection cycle 10 minutes / time, continuous for 24 hours;

[0081] Half-load operation: Equipment power 50%, fan speed 50%, data collection cycle 5 minutes / time, continuous for 24 hours;

[0082] Full load condition: Equipment power 100%, fan speed 100%, data collection cycle 2 minutes / time, continuous for 24 hours;

[0083] Baseline data processing:

[0084] Statistical characteristic calculation: Calculate the mean of each parameter. Standard deviation The distribution pattern was determined by the Shapiro-Wilk test (P>0.05 indicates a normal distribution).

[0085] Correlation analysis: Calculate the Pearson correlation coefficients among the parameters and establish a parameter correlation matrix;

[0086] Determining the normal fluctuation range: Based on the 95% confidence interval, the normal range of each parameter is determined: the positive and negative fluctuation range of temperature, humidity, and wind speed is twice the standard deviation of each parameter;

[0087] Model accuracy verification: Substitute the collected data into the calibrated model and calculate the model's calculated values. Compared with measured values Relative error: ,Require If the error is less than 5%, or if it exceeds the tolerance, recheck the sensor calibration accuracy or return to step S2 to adjust the model parameters until the accuracy requirements are met.

[0088] S402: Sensor Confidence Assessment System Initialization:

[0089] Set confidence level assessment metrics:

[0090] Consistency assessment: For the target sensor, select all neighboring sensors of the same type with a correlation coefficient R ≥ 0.7, and calculate the deviation rate: If the actual deviation rate is less than or equal to the deviation rate threshold, then the consistency score is... If the actual deviation rate is greater than the deviation rate threshold, the consistency assessment score will be lower. ;

[0091] It should be further noted that the threshold for temperature deviation rate is 3%, the threshold for humidity deviation rate is 5%, the threshold for wind speed deviation rate is 8%, and the threshold for vibration deviation rate is 10%.

[0092] Stability assessment: For the target sensor, continuous data is collected over one hour, and the fluctuation amplitude is calculated. The fluctuation amplitude is the maximum value measured by the target sensor within one hour minus the minimum value measured. A score is then calculated based on the fluctuation amplitude. If the actual fluctuation amplitude is less than or equal to the fluctuation threshold, the stability assessment score is [not specified]. If the actual fluctuation amplitude is greater than the fluctuation threshold, then ;

[0093] It should be further noted that the temperature fluctuation threshold is 0.5℃, the humidity fluctuation threshold is 2%RH, the wind speed fluctuation threshold is 0.3m / s, and the vibration fluctuation threshold is 0.05m / s².

[0094] Correlation assessment: Based on baseline data or physical models, determine the theoretical correlation coefficient between the target parameter and the associated parameters. Collect at least 30 sets of synchronized data from the target sensor and the associated parameter sensor, and calculate the actual correlation coefficient. Calculate the correlation bias: If the correlation deviation is less than or equal to the correlation threshold, the correlation assessment score is... If the correlation deviation is greater than the correlation threshold, then The correlation bias threshold is 0.1.

[0095] It should be further noted that both the theoretical correlation coefficient and the actual correlation coefficient are calculated using the Pearson correlation coefficient formula;

[0096] Confidence Calculation: Sensor Confidence ;

[0097] Adaptive diagnostic threshold setting: Based on the percentile of 72-hour baseline data, three levels of thresholds are set:

[0098] Note the thresholds: take the 90th percentile of temperature, humidity and wind speed;

[0099] Warning threshold: 95th percentile of temperature, humidity, and wind speed;

[0100] Danger threshold: 99th percentile of temperature, humidity, and wind speed;

[0101] At the same time, a threshold self-learning mechanism is established: the percentile is updated every 30 days based on newly collected steady-state data, and the threshold is automatically adjusted when the deviation is ≥10%.

[0102] S5: Multi-sensor data fusion and dynamic calibration: Preprocessing and time alignment are performed on the acquired data, Kalman filtering is used for noise reduction, and data is fused based on evidence theory according to sensor confidence. Dynamic data calibration is achieved through a three-level correlation calibration mechanism, and the calibration results are fed back. The specific steps are as follows:

[0103] S501: Data Preprocessing

[0104] Outlier removal: The 3σ criterion is used to remove outliers. The samples are replaced with the valid values ​​from the previous time step;

[0105] Time alignment: Based on the time synchronization signal in step S302, the data collected by different sensors are aligned to the same timestamp, with a time deviation of ≤1ms;

[0106] Data standardization: standardizing each parameter according to... Standardization eliminates the influence of dimensions, among which... The data is standardized, and x represents the data collected in real time.

[0107] S502: Multi-sensor data fusion

[0108] Kalman filtering noise reduction: Perform Kalman filtering on the data from a single sensor and update the optimal estimate;

[0109] DS Evidence Theory Fusion: For multi-sensor filtered data at the same measurement point, a basic probability allocation function is determined based on the sensor confidence level C. The data is then fused using evidence combination rules, and the final fused value is output.

[0110] Spatiotemporal consistency verification: Based on the correlation matrix in step S102, verify the spatiotemporal consistency of the fused data from adjacent sensors. If there is inconsistency, trigger local re-acquisition.

[0111] S503: Dynamic Correlation Calibration

[0112] Level 1 calibration, single sensor and model comparison calibration: compare the initial fused value of the sensor with the theoretical value calculated by the thermal balance model, humidity diffusion model or airflow motion model constructed in step S2. If the relative deviation between the two is ≤5%, the fused value is directly output as valid data; if the deviation is >5%, proceed to Level 2 calibration.

[0113] Secondary calibration, neighboring sensor correlation interpolation calibration: Based on the sensor correlation matrix constructed in step S1, neighboring sensors with a comprehensive correlation degree ≥ 0.7 with the target sensor are selected, effective fusion data of the neighboring sensors are extracted, and the calibration value of the target sensor is calculated using the inverse distance weighting method; if there are no neighboring sensors that meet the requirements, proceed to the tertiary calibration.

[0114] Level 3 calibration, model-assisted calibration: Based on the real-time parameters of the current environment, the corresponding category of associated measurement model is called to recalculate the theoretical value, and the theoretical value is used as the calibration value of the target sensor, and the model-assisted calibration label is marked.

[0115] Calibration feedback update: Record the deviation value of each calibration level, substitute it into the sensor confidence evaluation system in step S4, and dynamically update the confidence value of the target sensor.

[0116] S6: Measurement and Control: Defines four measurement modes: normal, early warning, fault, and maintenance. Designs mode conversion logic based on multi-parameter joint criteria and allocates data acquisition resources through environmental risk assessment. The specific steps are as follows:

[0117] S601: Measurement Mode Definition and Conversion Logic: Define four types of measurement modes:

[0118] Normal mode: Full parameter acquisition, sampling rate 1 time / minute;

[0119] Early warning mode: Enhanced data collection, sampling rate 1 time / 10 seconds;

[0120] Fault mode: Full parameter high-frequency acquisition, sampling rate 1 time / second, recording detailed data such as vibration time-domain waveform, and disabling sleep mode;

[0121] Maintenance mode: Full parameter acquisition and sensor self-calibration, sampling rate 1 time / 30 seconds, calibration time less than 5 minutes;

[0122] Mode switching logic:

[0123] From normal to early warning: Meet any of the following conditions: temperature ≥ T1 for 5 minutes, humidity ≥ H1 and temperature ≤ 20℃, wind speed ≤ 0.5m / s and ΔT ≥ 5℃;

[0124] From warning to fault: Meet any of the following conditions: temperature ≥ T2 or ΔT ≥ 5℃ within 10 minutes, vibration amplitude ≥ 0.1mm, or deviation of 2 or more sensors of the same type ≥ 2℃ or 5%RH;

[0125] From failure to normal: All parameters returned to the normal range and remained so for 30 minutes;

[0126] Maintenance mode: Manually triggered or triggered weekly. After calibration, it will revert to the current mode.

[0127] S602: Dynamic allocation of measurement resources:

[0128] Environmental risk assessment: Calculate the risk value for each measurement point: ,in, For parameter weights, The current value, This is the normal upper limit. The threshold is the warning threshold, and m is the total number of environmental parameters participating in the risk assessment at this measurement point;

[0129] It should be further explained that, It refers to the real-time measurement value of the i-th environmental parameter at a certain measurement point after processing by S5; This refers to the maximum allowable value of the i-th parameter under normal operating conditions, that is, the critical upper limit of the parameter within the safe range. Based on step S401, by collecting 72 hours of data from the power distribution room under three typical operating conditions of no load, half load, and full load, the statistical characteristics of each parameter are calculated, and based on the 95% confidence interval or actual operation and maintenance experience, the upper limit of normal fluctuation in the benchmark data is taken as this value. The threshold value at which the i-th parameter reaches the state of risk that needs to be alerted is used as the initial value based on the 95th percentile of S402, and is subsequently updated through a threshold self-learning mechanism. This refers to the weight of the i-th parameter on the environmental risk of the distribution box room, reflecting the importance of this parameter in risk assessment. It is determined by combining the analytic hierarchy process with failure case statistics.

[0130] Risk Level Classification: Based on the risk value, the measurement area is divided into three risk levels: high-risk area... Medium-risk areas Low-risk areas ;

[0131] Resource allocation rules:

[0132] High-risk areas: Prioritize the allocation of data collection resources, activate backup sensors, and increase the sampling rate to fault mode;

[0133] Medium-risk areas: Sampling rate increased to early warning mode, with a focus on collecting risk parameters;

[0134] Low-risk areas: Maintain normal operating procedures and reduce the sampling rate of non-critical parameters;

[0135] S7: Comprehensive Measurement and Status Assessment and Early Warning: Extracts multi-dimensional status parameters such as heat, humidity, airflow, and physical environment; calculates comprehensive assessment values ​​using normalization processing, dynamic weight allocation, and combination algorithms; analyzes status trends; and executes early warning responses according to levels. The specific steps are as follows:

[0136] S701: Multi-dimensional state parameter extraction:

[0137] Thermal state parameters: absolute temperature, relative temperature rise, temperature distribution uniformity, thermal time constant, thermal stability;

[0138] Humidity status parameters: relative humidity, humidity gradient, condensation risk, humidity stability;

[0139] Airflow parameters: wind speed distribution uniformity, air pressure gradient, ventilation efficiency;

[0140] Physical environmental parameters: vibration intensity, noise level, PM2.5 concentration;

[0141] It should be further explained that among the thermal state parameters, absolute temperature is the measured temperature value of key points obtained by multi-sensor data fusion, relative temperature rise is the difference between the absolute temperature of the heating element surface and the ambient reference temperature in the middle of the computer room, temperature distribution uniformity is the ratio of the difference between the highest and lowest temperatures of all measurement points in a certain area to the average temperature of the area, thermal time constant is the time required for the heating element to rise from the initial temperature to 63.2% of the final steady-state temperature, and thermal stability is the ratio of the temperature fluctuation amplitude of a certain measurement point in 1 hour to the average temperature of that hour.

[0142] Among the humidity status parameters, relative humidity is the measured humidity value of key points obtained by multi-sensor data fusion; humidity gradient is the difference between the relative humidity of the measurement point at the top space of the computer room and the relative humidity of the measurement point at the bottom space; condensation risk is the difference between the actual temperature and the dew point temperature of the measurement point; and humidity stability is the ratio of the relative humidity fluctuation of a measurement point within 1 hour to the average humidity of that hour.

[0143] Among the airflow state parameters, wind speed distribution uniformity is the difference between the highest and lowest wind speeds at all measurement points along the ventilation path, relative to the average wind speed; air pressure gradient is the air pressure difference between the air inlet and outlet of the computer room; and ventilation efficiency is the percentage of heat carried away by the ventilation system to the total heat generated by the equipment.

[0144] S702: Comprehensive Condition Assessment Calculation:

[0145] Normalize each parameter:

[0146] Positive parameters: Negative parameter: ;

[0147] in, , The 5th and 95th percentiles of the baseline data. The constraints are between 0 and 1, where i is an ordinal number;

[0148] It should be further explained that the larger the value of the positive parameter, the better the environmental condition and the lower the risk. These parameters include: thermal time constant, condensation risk, air pressure gradient, and ventilation efficiency. The larger the value of the negative parameter, the worse the environmental condition and the higher the risk. These parameters include: absolute temperature, relative temperature rise, temperature distribution uniformity, thermal stability, relative humidity, humidity gradient, humidity stability, wind speed distribution uniformity, vibration intensity, noise level, and PM2.5 concentration.

[0149] Dynamic weight allocation:

[0150] Sensitivity weights: ,in, For the rate of change of the parameter, The standard deviation of the historical rate of change;

[0151] Basic weights: The influence weights are the same as those in step S602, and are determined based on the analytic hierarchy process.

[0152] Final weights: ;

[0153] Comprehensive evaluation calculation: m represents the total number of environmental parameters;

[0154] S703: Trend Forecasting and Early Warning Response

[0155] Trend Analysis: 1-Hour Short-Term Trend: Least Squares Method for Fitting a Linear Trend Line and Calculating the Slope k>0 indicates an improvement in the situation, and k<0 indicates a deterioration; 30-day long-term trend: Predicting the next 7 days based on the ARIMA time series model. ;

[0156] Status levels are classified based on calculated comprehensive evaluation values:

[0157] Excellent: 90-100 points: All parameters are normal, output a health report;

[0158] Good: 75-89 points: slight deviations in individual parameters, focus on the deviation parameters;

[0159] Note: 60-74 points: Switch to alert mode and receive SMS reminder;

[0160] Warning: 40-59 minutes: Switch to fault mode, push audible and visual alarms and maintenance work orders;

[0161] Danger: 0-39 points: Emergency notification to the person in charge; optional triggering of equipment power-off protection.

[0162] Early warning recording and tracing: Automatically records the triggering conditions, response measures and processing results of each early warning, forming an early warning ledger for model optimization and threshold adjustment.

[0163] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other.

[0164] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An integrated measurement method for environmental parameters of a power distribution box room, characterized in that, include: S1: Overall design of the measurement system: Perform three-dimensional topology planning for multi-parameter measurement points, select and adapt sensors and construct sensor correlation matrix based on the correspondence between measurement points and parameters; S2: Construction and parameter calibration of multivariate correlation measurement model: Establish thermal equilibrium model, humidity diffusion model and airflow motion model, build a calibration system through a controllable environment chamber, perform experiments to collect data by parameter division, fit model parameters and establish a correction function library; S3: Measurement System Deployment and Synchronization Calibration: Install sensors and measurement units, configure master-slave communication links, achieve full system time synchronization through hardware triggering and software protocols, build a cross-validation network and perform initial calibration; S4: Measurement benchmark establishment and self-verification system initialization: Execute 72 hours of continuous data acquisition under different working conditions, process the data to obtain parameter statistical characteristics, correlation matrix and normal fluctuation range, verify the model accuracy, establish a sensor confidence evaluation system and set adaptive diagnostic thresholds; S5: Multi-sensor data fusion and dynamic calibration: Preprocessing and time alignment of the acquired data, Kalman filtering for noise reduction, and evidence theory fusion of data based on sensor confidence. Dynamic calibration of data is achieved through a three-level correlation calibration mechanism and the calibration results are fed back. S6: Measurement and Control: Defines four types of measurement modes: normal, early warning, fault, and maintenance. Designs mode conversion logic based on multi-parameter joint criteria and realizes data acquisition resource allocation through environmental risk assessment. S7: Comprehensive Measurement Status Assessment and Early Warning: Extract multi-dimensional status parameters such as heat, humidity, airflow and physical environment, use normalization processing, dynamic weight allocation and combination algorithm to calculate comprehensive assessment value, analyze status trend and execute early warning response according to level.

2. The integrated measurement method for environmental parameters of a power distribution box room according to claim 1, characterized in that: The construction of the sensor correlation matrix includes: Thermal conductivity correlation: Based on thermodynamic principles, the thermal conductivity of the materials between the two sensor locations, the heat exchange area, the average temperature difference in historical monitoring data, and the straight-line distance in three-dimensional space are quantitatively calculated. The value range is limited to 0-1, and the larger the value, the stronger the thermal conductivity correlation between the two locations. Airflow path correlation: The airflow characteristics at the locations of the two sensors are analyzed by CFD simulation. If the airflow can directly cover the two locations, the correlation value is 0.8-1.

0. If the airflow covers the locations through an indirect path, the correlation value is 0.3-0.

7. If there is no airflow coverage, the correlation value is 0-0.

2. Physical distance correlation: Based on a 5-meter baseline distance, the correlation value is normalized according to the three-dimensional Euclidean distance between the two sensors. The closer the distance, the higher the correlation value. The value range is 0-1. The weighting coefficients for each correlation dimension were determined through a combination of expert scoring and CFD simulation verification. Among them, the weighting coefficients for heat conduction correlation accounted for 40%, airflow path correlation accounted for 30%, and physical distance correlation accounted for 30%. The correlation between each sensor and the other sensors in three dimensions is weighted according to their respective weights and then summed to obtain the comprehensive correlation between each pair of sensors. Based on all the comprehensive correlations, an n×n matrix is ​​constructed, where n is the total number of sensors. The value of each element in the matrix is ​​the comprehensive correlation between the corresponding two pairs of sensors.

3. The integrated measurement method for environmental parameters of a power distribution box room according to claim 2, characterized in that: S2 includes: constructing three types of correlation measurement models: Thermal balance model: characterizes the balance between heat accumulation, convection, and radiation heat dissipation terms of the heating element; Humidity diffusion model: describes the spatial diffusion law of humidity, including temperature-dependent diffusion coefficient, humidity gradient and the influence of humidity source term; Airflow motion model: Establish the nonlinear relationship between ventilation outlet wind speed and machine room temperature difference and fan speed; A calibration experimental system was built, and calibration experiments were carried out for each model: the operating conditions were adjusted according to temperature, humidity, and wind speed steps, and environmental parameters, equipment operating parameters, and measured data of the corresponding models were collected; the least squares method was used to fit the experimental data and determine the parameters of each model.

4. The integrated measurement method for environmental parameters of a power distribution box room according to claim 1, characterized in that: The verification of model accuracy includes: After system deployment and calibration, collect measured environmental parameter data under different operating conditions, substitute the measured data into the thermal balance model, humidity diffusion model and airflow motion model constructed in S2, and calculate the theoretical output value of the model; compare the relative error between the theoretical value of the model and the measured data, and require that the relative error of all core parameters be less than 5%; if the error exceeds the tolerance, recheck the sensor calibration accuracy or return to step S2 to adjust the model parameters until the accuracy requirements are met.

5. The integrated measurement method for environmental parameters of a power distribution box room according to claim 1, characterized in that: The establishment of a sensor confidence assessment system and the setting of adaptive diagnostic thresholds include: constructing a three-dimensional sensor confidence assessment system. Consistency assessment: For the target sensor, the deviation rate between the target sensor and the reference data is calculated using the data of neighboring sensors with a comprehensive correlation degree ≥ 0.7 as a reference. Based on the calculated deviation rate and the deviation rate threshold of this type of data, the consistency assessment score is calculated. If the calculated deviation rate is greater than the deviation rate threshold of this type of data, the consistency assessment score is 0. Stability assessment: Statistically measure the parameter fluctuation range of the target sensor within 1 hour and calculate the stability assessment score. If the collected fluctuation range is greater than the fluctuation threshold, the stability assessment score is 0. Correlation assessment: Based on benchmark data or physical models, determine the theoretical correlation coefficient between the target parameter and the associated parameter, collect at least 30 sets of synchronous data from the target sensor and the associated parameter sensor, calculate the actual correlation coefficient, calculate the correlation deviation, and calculate the correlation assessment score based on the calculated correlation deviation and the correlation deviation threshold. If the calculated correlation deviation is greater than the correlation deviation threshold, the correlation assessment score is 0. Confidence score calculation: The confidence score of the sensor is obtained by weighting the consistency assessment score (40%), the stability assessment score (30%), and the correlation assessment score (30%) by summing the three assessment scores. Threshold benchmark: Based on the percentile of 72-hour working condition benchmark data. Note that the threshold corresponds to the 90th percentile, the warning threshold corresponds to the 95th percentile, and the danger threshold corresponds to the 99th percentile. The historical database is updated weekly by collecting 24-hour steady-state operating data of the equipment. The percentiles of each parameter are recalculated every 30 days. If the new threshold deviates from the current threshold by ≥10%, the threshold is automatically adjusted.

6. The integrated measurement method for environmental parameters of a power distribution box room according to claim 1, characterized in that: The three-level correlation calibration mechanism includes: First, outlier removal and time alignment are performed on the collected data. Then, multi-sensor data fusion is completed through Kalman filtering noise reduction and DS evidence theory to obtain the initial fused values ​​of the sensors. Level 1 calibration, single sensor deviation calibration: compare the sensor fused value with the theoretical value calculated by the S2 model. If the deviation is ≤5%, output directly; if the deviation is greater than 5%, proceed to Level 2 calibration. Level 2 calibration, sensor group calibration: Based on the sensor correlation matrix, it calls the fused data of neighboring sensors with correlation R≥0.7 and calculates the current sensor calibration value through the inverse distance weighting method; if there are no neighboring sensors that meet the requirements, it proceeds to Level 3 calibration; Level 3 calibration, model-assisted calibration: Based on the real-time parameters of the current environment, the corresponding category of associated measurement model is called to recalculate the theoretical value, and the theoretical value is used as the calibration value of the target sensor, and the model-assisted calibration label is marked. Calibration result feedback: Record the deviation value of each calibration level, substitute it into the sensor confidence evaluation system, and update the confidence value of the target sensor.

7. The integrated measurement method for environmental parameters of a power distribution box room according to claim 1, characterized in that: The design mode switching logic includes: four types of measurement modes: normal mode, early warning mode, fault mode, and maintenance mode, which are triggered independently and can be restored to the current operating mode; The transition from normal mode to early warning mode is triggered when any of the following multi-parameter joint criteria are met: the temperature reaches the temperature attention threshold and lasts for 5 minutes, the humidity reaches the humidity attention threshold and the ambient temperature is ≤20℃, the ventilation outlet wind speed is ≤0.5m / s and the temperature difference between the inside and outside of the computer room is ≥5℃. The transition from warning mode to fault mode is triggered when any of the following criteria are met: temperature reaches the warning threshold, temperature change ≥ 5℃ within 10 minutes, vibration amplitude ≥ 0.1mm, or data deviation of 2 or more similar sensors ≥ 2℃ or 5%RH. Fault mode to normal mode transition: Automatic switching occurs when all environmental parameters return to normal fluctuation range and this stable state lasts for 30 minutes without recurrence. Maintenance mode triggering and recovery: It can be triggered manually or automatically at a fixed time every week to perform sensor self-calibration. After calibration, it will automatically recover to the operating mode before the trigger.

8. The integrated measurement method for environmental parameters of a power distribution box room according to claim 1, characterized in that: The method of allocating resources through environmental risk assessment includes: Environmental risk value assessment: Based on real-time environmental parameter data of each measurement point, combined with parameter weights, normal upper limits of parameters and warning thresholds, the environmental risk value of each measurement point is quantitatively calculated. The magnitude of the risk value is positively correlated with the degree to which the parameter deviates from the normal range. Risk level classification: The measurement area is divided into three risk levels according to the risk value: high-risk area, medium-risk area, and low-risk area; High-risk area: measurement resources are allocated first, backup sensors are automatically activated to build redundant acquisition channels, and the sampling rate is increased to the fault mode; Medium-risk area: the sampling rate is adjusted to the early warning mode; Low-risk area: the normal mode sampling rate is maintained.

9. The integrated measurement method for environmental parameters of a power distribution box room according to claim 1, characterized in that: The comprehensive evaluation value includes: Multi-dimensional parameter extraction: First, extract four types of core environmental state parameters, including thermal state parameters, humidity state parameters, airflow state parameters, and other physical environment parameters; Parameter normalization: Based on the 5th and 95th percentiles of historical baseline data, all parameters are divided into positive and negative parameters and normalized separately. Dynamic weight allocation: The weights are determined by the fusion of sensitivity weights and basic importance weights. The sensitivity weights are calculated based on the standard deviation of the real-time rate of change and the historical rate of change of the parameters, while the basic importance weights are determined by the analytic hierarchy process. The two weights are then normalized to obtain the final weights of each parameter. Combination algorithm calculation: First, the normalized scores of each parameter are geometrically averaged according to their corresponding weights. Then, the weighted arithmetic average of each parameter is calculated. The product of the two calculation results is used to obtain the preliminary comprehensive evaluation value. The final revised comprehensive evaluation value will be constrained to the range of 0-100 points.

Images