Anti-condensation and humidity control system for low voltage switchgear

Abstract

The application discloses a condensation-preventing and humidity-controlling system of a low-voltage switch cabinet, and relates to the technical field of environment adjusting and controlling of low-voltage electrical equipment, comprising: a data acquisition module which acquires data through a multi-zone temperature and humidity sensor; a humidity prediction module which, in combination with cabinet parameters and historical data, uses a long short-term memory neural network to predict 30-minute humidity and pre-evaluate condensation risk; a condensation-preventing execution module which starts PTC heating, axial flow ventilation or a semiconductor dehumidifying device according to the prediction result; a system linkage module which interacts with a power monitoring system and adjusts execution device parameters during high load; a parameter self-learning module which optimizes a prediction model and records optimal thresholds of the device under different environments; and a safety protection module which monitors heating temperature and condensate level and switches to an emergency strategy when a fault occurs. The application is comprehensive in monitoring, accurate in prediction, cooperative in execution device operation, adaptive to different environments and loads, reduces condensation faults, prolongs the service life of elements and improves the operation reliability of the switch cabinet.

Description

Technical Field

[0001] This invention relates to the field of environmental regulation and control technology for low-voltage electrical equipment, and more particularly to an anti-condensation and humidity control system for low-voltage switchgear. Background Technology

[0002] Low-voltage switchgear, as the core equipment of power distribution systems, is widely used in industrial plants, residential communities, commercial buildings, and other scenarios. The insulation performance of its internal busbars, circuit breakers, cables, and other electrical components directly determines the reliability of the power distribution system. Condensation is one of the main hidden dangers in the operation of low-voltage switchgear. When the humidity inside the cabinet rises to near saturation and there is a large temperature difference between the inside and outside of the cabinet, water vapor in the air will condense into liquid water on the surface of electrical components, causing a decrease in the insulation resistance of the components. This can lead to leakage or discharge, or even short circuit faults, or even burn out equipment and cause power outages. Traditional low-voltage switchgear anti-condensation measures mostly rely on single temperature and humidity monitoring and passive triggering control, which has significant technical shortcomings. On the one hand, the monitoring dimensions are extremely limited. Most systems only place temperature and humidity sensors in a single location inside the switchgear, collecting only the temperature and humidity data inside the cabinet, ignoring key parameters such as the temperature and humidity outside the cabinet, the temperature gradient on the cabinet surface, and air pressure. For example, in summer when the outdoor temperature is high and humidity is high, hot and humid air outside the cabinet can easily seep into the cabinet through ventilation gaps. If only the humidity inside the cabinet is monitored, the warning is often triggered only after a large amount of water vapor has accumulated, missing the best time to prevent condensation. Furthermore, the lack of a temperature gradient can lead to misjudgment. When the temperature difference between the upper and lower areas inside the cabinet exceeds 5°C, the convergence of the cold air at the bottom and the hot air at the top can easily form local condensation. Traditional single-point monitoring cannot capture such local risks.

[0003] Traditional systems suffer from insufficient humidity prediction capabilities and inadequate linkage between actuators, further exacerbating the risk of condensation. Existing technologies often employ fixed threshold triggering mechanisms, activating dehumidifiers or heaters only when the humidity inside the cabinet exceeds 85% RH (relative humidity). This reactive control cannot predict humidity trends in advance. If humidity rises rapidly within a short period, the cabinet may already be nearing condensation threshold when the system is activated, making it difficult to effectively suppress condensation. Furthermore, the lack of coordination logic between actuators means that heating, ventilation, and dehumidification devices often operate independently. For example, activating only the heating device can raise the cabinet temperature to reduce relative humidity, but overheating can cause the cabinet temperature to exceed the safe operating temperature of electrical components (e.g., the maximum allowable temperature for circuit breakers is 60°C), accelerating component aging. Activating only the ventilation device can introduce more moisture if the outside humidity is higher than the inside, exacerbating the humidity increase. When the dehumidifier operates alone, condensation and freezing can easily occur in low-temperature environments, leading to a sharp drop in dehumidification efficiency and even equipment damage. In addition, traditional systems do not take into account the differences in adaptability to different environments. For example, in high-altitude areas, the decrease in air pressure will cause changes in the air saturation humidity. If the humidity threshold of plain areas is still used for control, it is easy to "misjudge low humidity" or "miss the risk of high condensation", which will further reduce the anti-condensation effect.

[0004] Inadequate safety protection mechanisms and a lack of system linkage are also prominent problems in traditional anti-condensation systems. On the one hand, the safety monitoring of traditional systems is limited to the surface temperature of the heating device, failing to comprehensively monitor the condensate discharge of the dehumidifier, overload faults of the actuator, and the reliability of the sensors. For example, when the condensate collection box of the dehumidifier overflows, the accumulated water can easily drip onto electrical components, directly causing a short circuit. When a sensor malfunctions, the system, lacking redundancy, will fall into a "blind control" state, unable to determine the actual humidity inside the cabinet, leading to the failure of anti-condensation measures. On the other hand, traditional systems are completely independent of the power monitoring system of the low-voltage switchgear, failing to consider the correlation between the operating load of electrical components and humidity: when electrical components are operating under high load, the components themselves generate a large amount of heat. If the humidity inside the cabinet is high at this time, the combined effect of heat and moisture will accelerate the aging of insulation materials and may even cause insulation breakdown. However, traditional systems cannot obtain load data and continue to operate according to conventional strategies, resulting in a significant increase in the risk of condensation under high load conditions. Meanwhile, traditional systems lack parameter self-learning capabilities, and the temperature and humidity variation patterns differ significantly across seasons and regions. Maintenance personnel must manually adjust control thresholds, increasing labor costs and potentially causing fluctuations in anti-condensation effectiveness due to untimely adjustments. These issues collectively contribute to the persistently high condensation failure rate of traditional low-voltage switchgear, severely impacting the stable operation of power distribution systems. Therefore, a more comprehensive, precise, and reliable anti-condensation and humidity control system is urgently needed. Summary of the Invention

[0005] The present invention proposes an anti-condensation and humidity control system for low-voltage switchgear to solve the problems mentioned in the prior art.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: an anti-condensation and humidity control system for a low-voltage switchgear, comprising: The data acquisition module collects real-time humidity H, internal temperature T, external temperature T0, external humidity H0, and temperature gradient change rate k on the surface of the low-voltage switchgear. The temperature gradient change rate k is calculated by the ratio of the temperature difference between two adjacent temperature sensors to the installation distance. The humidity prediction module, based on the collected data, combined with the structural parameters L and ventilation porosity ρ of the switchgear, as well as historical operating data, uses a long short-term memory neural network algorithm to predict the humidity change trend inside the cabinet within the next 30 minutes and outputs a predicted humidity curve; when the humidity inside the cabinet is in the range of 40%RH-70%RH, a condensation risk pre-assessment is initiated and the results are generated. The anti-condensation module selectively activates heating, ventilation, or dehumidification devices based on pre-assessment results; when the predicted humidity exceeds 85%RH and lasts for more than 10 minutes, the heating device is activated first. The system linkage module is used to interact with the power monitoring system of the low-voltage switchgear to obtain the operating load P of the electrical components inside the switchgear. l When the load on electrical components exceeds 80% of the rated load, the operating parameters of the anti-condensation actuator are adjusted in conjunction with the load. During peak load periods, if the heating device is running, the heating power is increased; if the ventilation device is running, the fan speed is increased. The parameter self-learning module iteratively optimizes the neural network model parameters of the humidity prediction module based on historical operating data and actual anti-condensation effects, while recording the optimal start-up threshold and operating parameters of each actuator under different seasons and environments. The safety protection module monitors the surface temperature T of the heating device in real time. h When T h When the temperature exceeds 70°C, the heating power is forcibly reduced to 50W and an early warning is issued; the water level h in the condensate collection box of the dehumidifier is monitored, and when h exceeds 80% of the total capacity, the drain pump is started and maintenance is prompted; at the same time, in the event of a communication failure or abnormality of the actuator, the system automatically switches to the preset emergency anti-condensation strategy.

[0007] Furthermore, it also includes a condensation risk quantification assessment unit, which determines the likelihood of condensation occurring by calculating a condensation risk coefficient C, using the formula: C=(HH) s )×(T0-T) / (T0+273.15), where H is the real-time humidity inside the cabinet. sThe saturated humidity is defined as the humidity at the current cabinet temperature T, where T0 is the outside temperature and T is the inside temperature. When C is greater than 0.8, it is considered a high risk of condensation. The system will immediately start the heating device and increase its power to 250W, while simultaneously starting the ventilation device at a speed of 2000r / min and continuously monitoring the humidity changes inside the cabinet, updating the condensation risk coefficient every 5 seconds. If C is between 0.5 and 0.8, it is considered a medium risk of condensation. The system will prioritize starting the dehumidification device and calculate the C value every 10 seconds. If C is less than 0.5, it is considered a low risk of condensation. Only the ventilation device will be started for intermittent ventilation at a speed of 1500r / min.

[0008] Furthermore, when predicting humidity change trends, the humidity prediction module also introduces a ventilation efficiency correction factor η, which is calculated as: η=(v×ρ×A) / (L×S), where v is the air velocity outside the cabinet, ρ is the ventilation porosity, A is the area of ​​a single ventilation hole, L is the effective ventilation length of the cabinet, and S is the ventilation side area of ​​the cabinet. When the outdoor wind force is level 3-4, the air velocity v outside the cabinet will fluctuate between 0.5-3m / s. At this time, the v value is collected in real time by the wind speed sensor and substituted into the calculation of η. Then, η is weighted and fused with the output value of the intermediate layer of the long short-term memory neural network. The weight coefficient α is obtained by training based on the correlation between historical wind speed and prediction error.

[0009] Furthermore, in the anti-condensation execution module, the heating power P of the heating device is related to the rate of increase in humidity inside the cabinet, v. h There are multiple levels of linkage between them, when the humidity rises at a rate v h When the RH / min exceeds 0.5%, the heating power P should be adjusted according to P = 100 + 200 × (v h / 1.5), where v h This represents the real-time humidity rise rate; when v h Within the range of 0.5-1%RH / min, P gradually increased from 150W to 233W; when v h Within the range of 1-1.5%RH / min, P increases from 233W to 300W; if v h Exceeding 1.5%RH / min.

[0010] Furthermore, the data acquisition module will also collect the air pressure value inside the switch cabinet. The humidity prediction module combines air pressure values ​​when predicting humidity. For saturated humidity Make corrections, the corrected version ,in Standard atmospheric pressure This refers to the real-time air pressure inside the cabinet; in areas with an altitude of 1500m, the standard atmospheric pressure is approximately 84.5kPa. If the real-time air pressure inside the cabinet... At a pressure of 84 kPa and a current cabinet temperature T of 25°C, the uncorrected saturated humidity is... It is 23.04 g / m³, the corrected =23.04×(101.325 / 84)≈23.04×1.206≈27.8g / m³.

[0011] Furthermore, when the system linkage module interacts with the power monitoring system, it obtains the operating load P of the electrical components. l Surface temperature T e Circuit breaker opening and closing status, bus current value I m When the circuit breaker is in the closed state and the bus current I m If the anti-condensation actuator is operating the heating device when the current exceeds 70% of the rated current, the system will adjust the upper limit of the heating power from 300W to 250W, and adjust the power according to T. e The numerical values ​​are dynamically adjusted. In addition, when the power monitoring system detects an overload warning for a certain section of the bus, the system linkage module will adjust the anti-condensation strategy in advance, and reduce the humidity inside the cabinet to below 60%RH 5 minutes before the overload occurs.

[0012] Furthermore, when optimizing the long short-term memory neural network model of the humidity prediction module, the parameter self-learning module adopts a hierarchical learning strategy: first, the historical data is classified by season, and for the data of each season, the number of hidden layer neurons N and the learning rate η of the neural network are optimized separately. l The root mean square error (RMSE) is used as the evaluation index for model accuracy, and the calculation method is as follows: H ip Let i be the humidity value predicted in the i-th time. Let be the actual humidity value of the i-th time, and n be the number of samples; when the RMSE of a certain season is greater than 5%RH, the model for that season is re-optimized.

[0013] Furthermore, the safety protection module monitors the surface temperature T of the heating device. h 1. Water level h in the condensate collection box of the dehumidifier; 2. Operating current I of the heating device. h Voltage U h Ventilation device fan speed n x Dehumidifier operating power P c， When heating device I h With rated current I hn If the ratio exceeds 1.2 and persists for 5 seconds, it is determined to be an overload fault. The system first reduces the heating power to 50% of the rated power and monitors I. h 10 seconds later I h If the current still exceeds 1.1 times the rated current, the power supply will be cut off, and emergency ventilation and dehumidification strategies will be activated; when the ventilation device n xWith the set speed n xs If the deviation exceeds ±20% and lasts for 10 seconds, it is determined to be a fan failure. The system should switch to the backup ventilation duct or increase the dehumidifier power to 80%. If the dehumidifier P... c With rated power P cn If the ratio is less than 0.3 and the humidity inside the cabinet increases, it is determined that the dehumidification device has failed, and the system immediately starts the heating and ventilation device coordinated mode; the safety protection module divides the fault type into three levels: warning, fault, and emergency fault, which correspond to different audible and visual alarm signals and remote notification strategies.

[0014] Furthermore, the dehumidification device operation control in the anti-condensation execution module adopts a multi-parameter coordinated adjustment strategy based on humidity, temperature, and air pressure: dehumidification power P c The adjustment formula is P c =P cn ×[0.6+0.2×(H / 100)+0.1×(T / 30)+0.1×(P a0 / P a )], where P cn H is the rated power of the dehumidifier, T is the real-time humidity inside the cabinet, and P is the real-time temperature inside the cabinet. a The real-time air pressure inside the cabinet is used; the real-time calculation of the condensate discharge rate V is based on the formula V=0.05×t×(H-40), where t is the dehumidification time and H is the humidity inside the cabinet when dehumidification is started; the coordinated switching logic between the dehumidification device and the heating and ventilation devices is as follows: when the temperature inside the cabinet T is below 15℃, the heating device is started first instead of the dehumidification device to avoid frost formation inside the cabinet due to dehumidification at low temperatures; when the ventilation device is running, if the humidity outside the cabinet H0 is lower than the humidity inside the cabinet H and the difference between the temperature outside the cabinet T0 and the temperature inside the cabinet T is less than 5℃, the running time of the ventilation device is increased.

[0015] Furthermore, the temperature and humidity sensors in the data acquisition module adopt a redundant "master-slave" arrangement, with one master sensor and at least one slave sensor set up for each monitoring area. When the difference between the measured values ​​of the master and slave sensors exceeds 5%RH or 2℃, the sensor fault diagnosis program is automatically started. This program uses a Kalman filter algorithm to fuse the data from the master and slave sensors and calculate the fused data. ,in For the fused data at time k, For the fused data at time k-1, For Kalman gain, This is the current sensor measurement value. The observation matrix is ​​used; if the deviation between the fused data and the main sensor data still exceeds 3%RH or 1℃, the main sensor is determined to be faulty, and the system automatically switches to the slave sensor for data acquisition. The main sensor is then marked as faulty, and a sensor fault report is generated through the edge computing unit and pushed to the operation and maintenance platform.

[0016] Compared with existing technologies, the beneficial effects of this invention are: In terms of monitoring dimensions, the system uses multiple sensors to cover key areas such as the switchgear busbar room, cable room, and circuit breaker room. It also collects the temperature and humidity inside and outside the cabinet, as well as the temperature gradient of the cabinet. Compared with traditional single-point monitoring, the data collection is more comprehensive and can accurately capture the risk of condensation caused by local temperature and humidity differences. This avoids misjudgment of risks due to monitoring blind spots and provides a reliable data foundation for subsequent anti-condensation control.

[0017] The system's advantages are particularly evident in humidity prediction and control. The traditional "post-event triggering" mode is replaced by accurate prediction 30 minutes in advance. Combined with long short-term memory neural networks and correction factors such as ventilation efficiency and air pressure, the prediction results can adapt to different wind speeds and altitudes, effectively avoiding the problem of delayed response when humidity rises sharply. The actuators no longer work independently; the heating, ventilation, and dehumidification devices operate in coordination based on the prediction results and real-time operating conditions. For example, dehumidification and ventilation are prioritized when humidity is high and load is low, while heating power is adjusted and ventilation is linked to heat dissipation when humidity is high and load is high. This ensures dehumidification effect while avoiding overheating of components or energy waste, achieving a balance between condensation prevention and equipment safety.

[0018] The integrated design of the system with the power monitoring system further reduces the risk of condensation under high load conditions. By acquiring the operating load, surface temperature, and opening / closing status of electrical components, the system can dynamically adjust anti-condensation strategies—for example, reducing the humidity inside the cabinet in advance during peak load periods to avoid the combined effect of component heating and high humidity accelerating insulation aging. Compared with the traditional system's "out-of-condition" control method, this system better meets the actual operating needs of low-voltage switchgear and significantly improves operational safety during high load periods.

[0019] In terms of safety protection and adaptability, the system design is equally comprehensive. Faults such as heating device overload and condensate overflow can be monitored in real time and trigger tiered protection. The master-slave redundant arrangement of sensors avoids system paralysis caused by a single sensor failure. The parameter self-learning module can automatically optimize the prediction model and execution thresholds based on the environmental characteristics of different seasons and regions, reducing manual intervention and lowering operation and maintenance costs. Furthermore, the emergency anti-condensation strategy ensures that the switchgear can maintain a safe humidity environment even when communication or actuator malfunctions, further enhancing system reliability.

[0020] Overall, this invention significantly reduces the condensation failure rate of low-voltage switchgear through multi-dimensional monitoring, accurate prediction, collaborative execution, operating condition linkage, and intelligent protection. It also extends the service life of electrical components, reduces the workload of maintenance personnel and the cost of fault repair in the power distribution system, and provides strong support for the stable operation of low-voltage switchgear in complex environments and changing operating conditions. Attached Figure Description

[0021] Figure 1 This is a schematic block diagram of the anti-condensation and humidity control system for low-voltage switchgear proposed in this invention. Figure 2 Line graph comparing the 30-minute error of humidity prediction at different altitudes; Figure 3 A line graph showing the temperature change trend inside the cabinet under high-load conditions in summer; Figure 4 Grouped bar charts comparing the annual energy consumption of anti-condensation actuators in different seasons; Figure 5 A line graph showing the deviation of humidity data inside the cabinet when the sensor fails. Detailed Implementation

[0022] 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.

[0023] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0024] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. Furthermore, the terms "installed," "connected," and "linked" should be interpreted broadly; for example, they may refer to a fixed connection, a detachable connection, or an integral connection; they may refer to a mechanical connection or an electrical connection; they may refer to a direct connection or an indirect connection through an intermediate medium; and they may refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances. The invention will now be described in further detail with reference to the accompanying drawings.

[0025] Reference Figures 1 to 5 A low-voltage switchgear anti-condensation and humidity control system, comprising: The data acquisition module is used to collect real-time humidity H, internal temperature T, external temperature T0, external humidity H0, and temperature gradient change rate k on the surface of the low-voltage switchgear. The data acquisition module includes multiple temperature and humidity sensors, which are respectively arranged at different heights in the busbar compartment, cable compartment, and circuit breaker compartment of the switchgear. The temperature gradient change rate k is calculated by the ratio of the temperature difference between two adjacent temperature sensors to the installation distance. The sampling frequency is 5Hz to realize dynamic monitoring of the internal temperature and humidity and temperature gradient. The humidity prediction module, based on real-time data acquired by the data acquisition module, combined with the switch cabinet's structural parameters L (effective ventilation length of the cabinet), ventilation porosity ρ (ratio of total pore area to cabinet side area), and historical operating data, uses a long short-term memory neural network algorithm to predict the humidity change trend inside the cabinet within the next 30 minutes and outputs a predicted humidity curve; at the same time, when the humidity inside the cabinet is in the range of 40%RH-70%RH, a condensation risk pre-assessment is initiated and a pre-assessment result is generated; The anti-condensation execution module selectively activates the heating device, ventilation device, or dehumidification device based on the prediction and pre-assessment results of the humidity prediction module. The heating device is a PTC heater distributed at the bottom and back of the switch cabinet, the ventilation device is an axial flow fan with a dust filter, and the dehumidification device is a semiconductor condensation dehumidifier. When the predicted humidity will exceed 85%RH and the duration exceeds 10 minutes, the heating device is activated first, and the heating power P is gradually adjusted from 100W to 300W in a gradient. The system linkage module is used to interact with the power monitoring system of the low-voltage switchgear to obtain the operating load P of the electrical components inside the switchgear. l(Active power consumption per unit time); When the load on electrical components exceeds 80% of the rated load, the working parameters of the anti-condensation execution module are adjusted in conjunction with the load. During peak load periods, if the heating device is running, the heating power is increased appropriately; if the ventilation device is running, the fan speed is increased. The parameter self-learning module iteratively optimizes the parameters of the neural network model of the humidity prediction module based on historical operating data and actual anti-condensation effects. At the same time, it records the optimal start-up threshold and operating parameters of each actuator in different seasons and environments. For example, in the high temperature and high humidity environment of summer, the optimal start-up humidity threshold of the ventilation device is learned to be 65%RH, while in winter it is adjusted to 70%RH. The safety protection module monitors the surface temperature T of the heating device in real time. h When T h When the temperature exceeds 70°C, the heating power is forcibly reduced to 50W and an early warning is issued; the water level h in the condensate collection box of the dehumidifier is monitored, and when h exceeds 80% of the total capacity, the drain pump is started and maintenance is prompted; at the same time, in the event of a communication failure or abnormality of the actuator, the system automatically switches to the preset emergency anti-condensation strategy to ensure that the humidity in the switch cabinet does not exceed the limit.

[0026] This invention also includes a condensation risk quantification assessment unit, which determines the likelihood of condensation occurring by calculating a condensation risk coefficient C. The condensation risk coefficient C is calculated as follows: C = (HH) s )×(T0-T) / (T0+273.15), where H is the real-time humidity inside the cabinet. s The saturated humidity is the humidity at the current cabinet temperature T, where T0 is the outside temperature and T is the inside temperature. When C is greater than 0.8, it is considered a high risk of condensation. The system will immediately start the heating device and increase the power to 250W, while simultaneously starting the ventilation device at a speed of 2000r / min and continuously monitoring the humidity change inside the cabinet, updating the condensation risk coefficient every 5 seconds. If C is between 0.5 and 0.8, it is considered a medium risk of condensation. The system will prioritize starting the dehumidifier, operating at 50% of its rated power, and calculating the C value every 10 seconds. If C is less than 0.5, it is considered a low risk of condensation. Only the ventilation device will be started at a speed of 1500r / min for intermittent ventilation (running for 3 minutes and stopping for 1 minute). In addition, the unit will also take into account the operating time t0 of the switch cabinet (accumulated from the last condensation failure). When t0 exceeds 72 hours and C is greater than 0.6, even if it is in the medium risk range, the heating device will be activated to assist dehumidification with a power of 150W to avoid long-term low risk accumulation leading to sudden condensation.

[0027] In this invention, the humidity prediction module introduces a ventilation efficiency correction factor η when predicting humidity change trends. The ventilation efficiency correction factor η is calculated as follows: η=(v×ρ×A) / (L×S), where v is the air velocity outside the cabinet, ρ is the ventilation porosity, A is the area of ​​a single ventilation hole, L is the effective ventilation length of the cabinet, and S is the ventilation side area of ​​the cabinet. When the outdoor wind force is level 3-4, the air velocity v outside the cabinet will fluctuate between 0.5-3m / s. At this time, the v value is collected in real time by a wind speed sensor and substituted into the calculation of η. Then, η is weighted and fused with the output value of the intermediate layer of the long short-term memory neural network. The weight coefficient α is obtained by training based on the correlation between historical wind speed and prediction error (α is dynamically adjusted between 0.3 and 0.7). For example, when the external air velocity v is 2 m / s, the ventilation porosity ρ is 0.3, the area of ​​a single ventilation pore A is 2 cm², the effective ventilation length L of the cabinet is 1.2 m, and the ventilation side area S is 0.8 m², the calculated η is (2 × 0.3 × 2 × 10⁻⁶). -4 ) / (1.2×0.8)=1.2×10 -4 / 0.96≈0.000125, where α is 0.5. After fusion, the humidity prediction error can be reduced from 8%RH (uncorrected) to less than 5%RH. For switch cabinets with different structures (such as GGD and GCS types), their corresponding ρ, A, L, and S reference values ​​will be pre-stored and automatically called during system initialization to ensure the adaptability of η calculation.

[0028] In this invention, in the anti-condensation execution module, the heating power P of the heating device is related to the humidity rise rate v inside the cabinet. h There are multiple levels of linkage between them, when the humidity rises at a rate v h When the RH / min exceeds 0.5%, the heating power P should be adjusted according to P = 100 + 200 × (v h / 1.5), where v h The real-time humidity rise rate (unit: %RH / min); when v h Within the range of 0.5-1%RH / min, P gradually increased from 150W to 233W; when v h Within the range of 1-1.5%RH / min, P increases from 233W to 300W; if v h If the RH / min exceeds 1.5%, in addition to adjusting the power according to the formula above, the system will also trigger temperature feedback adjustment of the heating device, monitoring the cabinet temperature T in real time. If the rate of increase of T within 1 minute after power adjustment exceeds 1℃ / min, then P will be adjusted back to 80% of the current value to prevent excessively high cabinet temperature from affecting electrical components. For example, if v is monitored in a certain instance... hWith a RH / min of 1.2%, substituting into the formula, we get P = 100 + 200 × (1.2 / 1.5) = 100 + 160 = 260W. If the temperature inside the cabinet rises from 25℃ to 27℃ within 1 minute (rising rate 2℃ / min), then P will be adjusted back to 260 × 0.8 = 208W. At the same time, the ventilation device will be activated at a speed of 1000 r / min to assist in heat dissipation, thus achieving a dynamic balance between heating and temperature control.

[0029] In this invention, the data acquisition module also collects the air pressure value inside the switch cabinet. The humidity prediction module combines air pressure values ​​when predicting humidity. For saturated humidity Corrections were made, and the corrected saturation humidity was determined. ,in Standard atmospheric pressure (101.325 kPa). This refers to the real-time air pressure inside the cabinet; in areas with an altitude of 1500m, the standard atmospheric pressure is approximately 84.5kPa. If the real-time air pressure inside the cabinet... At a pressure of 84 kPa and a current cabinet temperature T of 25°C, the uncorrected saturated humidity is... It is 23.04 g / m³, the corrected =23.04×(101.325 / 84)≈23.04×1.206≈27.8g / m³. With this correction, when the real-time humidity H inside the cabinet is 25g / m³, the uncorrected H and... The initial difference was -1.96 g / m³ (easily misinterpreted as low humidity), while the corrected difference is -2.8 g / m³, which is closer to the actual saturation state. This improves the accuracy of humidity prediction in high-altitude areas by more than 20% compared to the uncorrected result. Simultaneously, the humidity prediction module will use the corrected value... Compared with the external humidity H0 and air pressure difference ( - Multi-parameter fusion is performed to further optimize the input dimension of the prediction model.

[0030] In this invention, when the system linkage module interacts with the power monitoring system, it acquires the operating load P of the electrical components. l and surface temperature T e In addition, it will also acquire the circuit breaker's opening and closing status and the bus current value I. m Operating parameters; when the circuit breaker is in the closed state and the bus current I m If the anti-condensation actuator is operating the heating device when the current exceeds 70% of the rated current, the system will adjust the upper limit of the heating power from 300W to 250W, and adjust the power according to T. e The value of T is dynamically adjusted: if T e Between 50-60℃, the heating power is calculated as P=250-5×(T)e Adjustments are made for -50), for example, T e At 55℃, P = 250 - 5 × 5 = 225 W; if T e If the temperature exceeds 60℃, the heating device will be immediately stopped, the ventilation device will be activated first, and the fan speed will be increased to 2500 r / min. At the same time, the dehumidifier will be turned on and run at 70% of its rated power. In addition, when the power monitoring system detects an overload warning for a certain bus section (lasting more than 3 minutes), the system linkage module will adjust the anti-condensation strategy in advance, reducing the humidity inside the cabinet to below 60%RH 5 minutes before the overload occurs, to avoid the risk of condensation discharge caused by the combined effects of component heating due to overload and high humidity.

[0031] In this invention, the parameter self-learning module employs a hierarchical learning strategy when optimizing the long short-term memory neural network model of the humidity prediction module: firstly, historical data is categorized by season (spring (March-May), summer (June-August), autumn (September-November), and winter (December-February); then, for each season's data, the number of hidden layer neurons N and the learning rate η of the neural network are optimized separately. l For example, in summer data samples, due to drastic temperature and humidity changes, N was adjusted from the default 64 to 96, and η... l The value was adjusted from 0.001 to 0.0015 to accelerate model convergence. Simultaneously, the root mean square error (RMSE) was adopted as the evaluation metric for model accuracy. The RMSE calculation method is as follows: ,in Let i be the humidity value predicted in the i-th time. Let be the actual humidity value of the i-th time, and n be the number of samples. When the RMSE of a certain season is greater than 5%RH, the model for that season is re-optimized. During the optimization process, a particle swarm optimization algorithm is introduced to perform a global search on the weight parameters of the neural network, and then gradient descent is used for local fine-tuning. After optimization, the RMSE of the summer model can be reduced from 7%RH to below 4%RH, significantly improving the seasonal adaptability of humidity prediction.

[0032] In this invention, the safety protection module monitors the surface temperature T of the heating device. h In addition to monitoring the water level h in the dehumidifier's condensate collection box, the operating current I of the heating device will also be monitored in real time. h and voltage U h The fan speed n of the ventilation device x The operating power P of the dehumidifier c When the operating current I of the heating device h With rated current I hn If the ratio exceeds 1.2 and lasts for more than 5 seconds, it is determined to be an overload fault of the heating device. The system first reduces the heating power to 50% of the rated power and continuously monitors I. h If I...h If the current still exceeds 1.1 times the rated current, disconnect the power supply to the heating device and activate the emergency ventilation and dehumidification strategy; when the fan speed of the ventilation device reaches n... x With the set speed n xs If the deviation exceeds ±20% and lasts for 10 seconds, it is determined to be a fan failure, and the system switches to the backup ventilation duct (if available) or increases the dehumidifier power to 80%; for the dehumidifier, if the operating power P c With rated power P cn If the ratio is less than 0.3 and the humidity inside the cabinet is still rising, it is determined that the dehumidifier has failed. The system immediately activates the coordinated operation mode of the heating and ventilation devices, setting the heating power to 200W and the fan speed to 2200r / min. In addition, the safety protection module classifies the fault type into three levels: warning (e.g., water level approaching 80%), fault (e.g., fan speed deviation), and emergency fault (e.g., heating device overload), each corresponding to different audible and visual alarm signals and remote notification strategies to ensure timely response by maintenance personnel.

[0033] In this invention, the operation control of the dehumidification device in the anti-condensation execution module adopts a multi-parameter coordinated adjustment strategy based on humidity, temperature, and air pressure: dehumidification power P c The adjustment formula is P c =P cn ×[0.6+0.2×(H / 100)+0.1×(T / 30)+0.1×( / )], where P cn H represents the rated power of the dehumidifier, H represents the real-time humidity inside the cabinet, and T represents the real-time temperature inside the cabinet. This refers to the real-time air pressure inside the cabinet; for example, when H is 75%RH and T is 28℃. At 90 kPa, P c =P cn ×[0.6+0.2×0.75+0.1×(28 / 30)+0.1×(101.325 / 90)]≈P cn ×(0.6+0.15+0.093+0.1126)≈0.9556×P cnThis means the dehumidifier operates at approximately 95.6% of its rated power. Simultaneously, the real-time calculation of condensate discharge volume V uses the formula V = 0.05 × t × (H - 40), where t is the dehumidification time (in minutes) and H is the humidity inside the cabinet when dehumidification starts (in %RH). When the calculated V exceeds 70% of the total capacity of the condensate collection box, the drain pump is started in advance for intermittent drainage (running for 1 minute, stopping for 2 minutes). Furthermore, the coordinated switching logic between the dehumidifier and the heating / ventilation devices is as follows: when the cabinet temperature T is below 15℃, the heating device is activated first, rather than the dehumidifier, to avoid frost formation inside the cabinet due to low-temperature dehumidification; when the ventilation device is running, if the external humidity H0 is lower than the internal humidity H and the difference between the external temperature T0 and the internal temperature T is less than 5℃, the running time of the ventilation device is increased to reduce the energy consumption of the dehumidifier.

[0034] In this invention, the temperature and humidity sensors of the data acquisition module adopt a "master-slave" redundant arrangement. Each monitoring area (busbar room, cable room, circuit breaker room) is equipped with one master sensor and at least one slave sensor. The master sensor is responsible for real-time data acquisition and uploading, while the slave sensors serve as backups. When the difference between the measured values ​​of the master and slave sensors exceeds 5%RH or 2℃, the sensor fault diagnosis program is automatically started. This program uses a Kalman filter algorithm to fuse the data from the master and slave sensors and calculate the fused data. ,in For the fused data at time k, For the fused data at time k-1, For Kalman gain, This is the current sensor measurement value. The system uses an observation matrix. If the fused data deviates from the main sensor data by more than 3%RH or 1℃, the main sensor is considered faulty. The system automatically switches to the slave sensor for data acquisition, marks the main sensor as faulty, and generates a sensor fault report via the edge computing unit, pushing it to the maintenance platform. Furthermore, the sensor installation locations are optimized using fluid dynamics simulation, placing them in areas with relatively stable airflow within the cabinet (such as the airflow outlet side of the busbar compartment or above the cable interlayer in the cable compartment) to ensure the representativeness of the collected data and reduce measurement errors caused by airflow disturbances.

[0035] The following two examples further illustrate the specific implementation of this system: Example 1: Low-voltage switchgear for power distribution room in high-humidity, high-altitude mountainous area (altitude 1800m, diurnal temperature difference 15℃, annual average humidity 75%RH) 1. Detailed configuration and operation process of system modules (1) Data acquisition module The power distribution room consists of GGD type low-voltage switchgear (3 units operating in parallel, including a busbar compartment, cable compartment, and circuit breaker compartment). The data acquisition module adopts a "master-slave redundancy + multi-area coverage" design: each cabinet has one set of master-slave temperature and humidity sensors (model SHT30, accuracy ±2%RH, ±0.3℃) in the busbar compartment (top 1 / 3 height) and one set of composite sensors with air pressure detection (model BME280, air pressure accuracy ±1hPa) in the middle of the circuit breaker compartment; wind speed sensors (model FS300, range 0-30m / s) and temperature and humidity sensors are installed on the outer wall of the cabinet. The sampling frequency is 5Hz, and the data is transmitted to the control cabinet (PLC model S7-1200) via RS485 bus.

[0036] Specific data collection parameters: Real-time humidity inside the cabinet H=72%RH, temperature inside the cabinet T=22℃, outside temperature T0=18℃, outside humidity H0=80%RH, air pressure inside the cabinet. =82kPa (standard atmospheric pressure) =101.325kPa), temperature gradient k=(busbar compartment 24℃-cable compartment 20℃) / 0.8m (distance between two sensors)=5℃ / m, wind speed outside the cabinet v=1.2m / s.

[0037] (2) Humidity prediction module (including formula application) Saturation humidity correction: At the current T=22℃, the saturation humidity is not corrected. =19.37g / m³, according to the formula calculate: =19.37×(101.325 / 82)≈19.37×1.236≈24.04g / m³, which is more consistent with the actual saturation state under high altitude and low air pressure environment.

[0038] Ventilation efficiency correction factor: Switchgear ventilation porosity ρ=0.25 (side panel with φ8mm round holes, total area 0.02m²), single ventilation hole area A=5.02×10 -5 m², effective ventilation length of the cabinet L = 1.5m, ventilation side area S = 0.6m², calculated using the formula η = (v × ρ × A) / (L × S): η = (1.2 × 0.25 × 5.02 × 10 -5 ) / (1.5×0.6)=(1.506×10 -5 ) / 0.9≈1.67×10 -5 The η is weighted and fused with the output of the LSTM neural network (weight α=0.6, due to the stable wind speed in the mountainous area) to predict the humidity change in the next 30 minutes: initial H=72%RH, 75%RH after 10 minutes, 78%RH after 20 minutes, and 81%RH after 30 minutes (not reaching 85%RH, but the risk of condensation needs to be assessed).

[0039] Condensation risk factor: According to the formula C=(H- )×(T0-T) / (T0+273.15), =19.37g / m³ (corresponding to saturated humidity at 22℃, which translates to approximately 90% RH relative humidity, hence H- (Relative humidity difference) = 72% - 90% = -18%, here it is corrected to relative humidity calculation: C = (72 - 90) × (18 - 22) / (18 + 273.15) = (-18) × (-4) / 291.15 ≈ 72 / 291.15 ≈ 0.247 < 0.5, judged as low condensation risk, but because the humidity at high altitude is prone to sudden rise, the system activates the ventilation device for pre-control.

[0040] (3) Anti-condensation implementation and system linkage Module actions: Under low condensation risk, the ventilation device (axial flow fan model 4E-230, rated speed 2000r / min) operates intermittently at 1800r / min (running for 4 minutes, stopping for 2 minutes); the dehumidification device (semiconductor dehumidifier model TE-03, rated power 80W) is on standby at 40% power to avoid low-temperature dehumidification and frost (the temperature inside the cabinet is 22℃, close to the frost threshold).

[0041] System linkage: The power distribution room is for rural power grid distribution, and the load fluctuates greatly during the irrigation season (May-September). When the power monitoring system feeds back the bus current I... m =450A (rated current 500A, load P) l When the load reaches 90% of rated load, the system will adjust accordingly: the ventilation fan speed will be increased to 2200 r / min, and the heating device (PTC heater model JRQ-300, rated power 300W) will be preheated to 150W for standby, to avoid the superposition of load heating and high humidity; when the load drops to 60% (I m When the current is 300A, the fan resumes to 1800r / min, and the heating device stops.

[0042] (4) Parameter self-learning and safety protection Self-learning optimization: During the rainy season (June-August) in mountainous areas, when humidity consistently exceeds 75% RH, the self-learning module records the optimal activation threshold for ventilation devices as 65% RH, and for dehumidifiers as 70% RH. During the dry season (December-February), with humidity between 50% and 60%, the optimization is to set the ventilation activation threshold to 70% RH, and to shut down the dehumidifiers. LSTM model seasonal optimization: During the rainy season, the hidden layer neurons N=96, and the learning rate η... l =0.0015; Dry season N=64, η l =0.001, and the root mean square error (RMSE) decreased from the initial 6.5%RH to 3.8%RH.

[0043] Safety protection: Surface temperature T of the heating deviceh Real-time monitoring (thermocouple type K), when misoperation causes T h At 73℃, the system is forced to reduce to 50W and triggers an audible and visual alarm (buzzer at 85dB, red light flashing); the water level h in the dehumidifier's condensate collection box is monitored by an infrared liquid level sensor. When h = 82% of the total capacity, the drain pump (model DC30A) starts draining water, and a maintenance reminder is pushed to the maintenance APP; in the event of a main sensor failure (humidity difference of 6%RH), the system automatically switches to the slave sensor, and the Kalman filter fusion data deviation is <2%RH, with no monitoring interruption.

[0044] 2. Performance Comparison Data Table 1: Comparison of Anti-condensation Effects of Low-Voltage Switchgear in High-Humidity and High-Altitude Mountainous Areas Humidity prediction error (30 minutes) ±8.2%RH ±3.8%RH <![CDATA[Air pressure correction + H s 'Fusion with η]]> Annual condensation failure rate 12.5% 1.8% Multi-regional monitoring + risk factor C Component overheating rate under high load conditions 8.3% 0.9% System-wide load regulation Sensor fault monitoring outage rate 7.1% 0% Master-slave redundancy + Kalman filtering Frequency of manual intervention in operation and maintenance 12 times / year 3 times / year Parameter self-learning + automatic maintenance Table 1 is a statistical summary of one year's operation data for this power distribution room. Traditional systems, due to the lack of correction for high-altitude saturated humidity, have a 30-minute humidity prediction error of ±8.2%RH, and are prone to misjudging low humidity during the rainy season, leading to condensation. The annual failure rate is 12.5%. This invention utilizes H... s By integrating correction and η, the error is reduced to ±3.8%RH, and the risk coefficient C is used to identify potential spikes under low-risk conditions in advance, with a condensation rate of only 1.8%. Under high load, traditional systems experience conflicts between heating and heat dissipation, resulting in an overheating rate of 8.3% for components; this invention dynamically adjusts power based on load, reducing the overheating rate to 0.9%. When sensors fail, traditional systems interrupt monitoring; this invention uses master-slave redundancy and filtering to ensure uninterrupted operation, reducing maintenance interventions from 12 times / year to 3 times, significantly lowering maintenance costs in mountainous areas and meeting the needs of unattended power distribution rooms in remote regions.

[0045] Example 2: Low-voltage switchgear in a high-temperature and high-humidity coastal industrial area (altitude 5m, average summer temperature T=32℃, H=85%RH, typhoon season wind speed 0-12m / s) 1. Detailed configuration and operation process of system modules (1) Data acquisition module The industrial area uses 10 GCS-type drawer-type low-voltage switchgear units (for three-shift production). Data acquisition modules are designed for high-density deployment: each drawer circuit breaker compartment and busbar compartment has one set of master-slave temperature and humidity sensors (SHT35, accuracy ±1.5%RH); wind speed and direction sensors (WindSonic, range 0-60m / s) are installed on the cabinet top; and temperature and humidity sensors (IP65 outdoor protection rating) are installed on the exterior walls. The sampling frequency is 5Hz, and data is transmitted to an industrial IoT platform via Ethernet. Typical acquisition parameters: During a summer typhoon, inside the cabinet H=82%RH, T=35℃; outside the cabinet T0=33℃, H0=88%RH, wind speed v=5.8m / s, and air pressure... =100.5kPa, bus current I m=580A (rated 630A, P) l =92% of rated load), temperature gradient k = (top 37℃ - bottom 33℃) / 1.2m ≈ 3.3℃ / m.

[0046] (2) Humidity prediction module Ventilation efficiency correction: Switchgear ventilation porosity ρ=0.3 (side panel louvers, total area 0.03m²), single pore A=8×10 -5 Given m², effective ventilation length L = 1.8m, and ventilation side area S = 1.0m², calculate using the formula η = (v × ρ × A) / (L × S): η = (5.8 × 0.3 × 8 × 10 -5 ) / (1.8×1.0)=(1.392×10 -4 ) / 1.8≈7.73×10 -5 With a weight α=0.7 (wind speed fluctuates greatly during typhoons, and η has a significant impact), LSTM predicts humidity for 30 minutes: initial 82%RH → 86%RH for 20 minutes → 88%RH for 30 minutes (exceeding the 85%RH threshold).

[0047] Condensation risk factor: 22℃ saturated humidity =19.37g / m³, at 35℃ =42.43g / m³ (relative humidity 100%), H- (Relative humidity difference) = 82% - 100% = -18%. According to the formula C = (82 - 100) × (33 - 35) / (33 + 273.15) = (-18) × (-2) / 306.15 ≈ 36 / 306.15 ≈ 0.118 < 0.5. However, the predicted humidity exceeds 85%RH, so the system starts the anti-condensation operation.

[0048] (3) Anti-condensation implementation and system linkage Module Actions: If the predicted humidity exceeds 85%RH, the dehumidifier (TE-05, rated 120W) will be activated first, operating at 80% power, and the ventilation fan (4E-300, rated 2500r / min) will operate at 2200r / min. After 15 minutes, when H=84%RH, the heating device (JRQ-300) will be activated at 150W (to avoid overheating; the internal temperature T≤40℃). After 25 minutes, when H=82%RH, the dehumidifier will be restored to 60% power + fan at 2000r / min.

[0049] System linkage: When the factory motor starts, I m =610A (P) l (At 97% rated load), upon feedback from the power monitoring system, the system immediately stopped the heating device, increased the fan speed to 2500 r / min, maintained the dehumidification power at 80%, and simultaneously monitored the circuit breaker surface temperature T. e=58℃ (≤60℃ safety value); load reduced to 70% (I m When the wind speed reaches 440A, restore the heating to 100W for standby. During a typhoon, if the wind speed suddenly increases to 10m / s, η is recalculated as 1.36×10⁻⁶. -4 The predicted humidity was corrected to 85%RH, and the system reduced the fan speed to 1800r / min (to prevent strong winds from blowing in hot and humid air).

[0050] (4) Parameter self-learning and safety protection Self-learning optimization: In summer (June-September) with high temperature and humidity, the self-learning ventilation start threshold is 65%RH, and dehumidification starts at 70%RH, with LSTMN=96 and RMSE=3.5%RH; in winter (December-February) with T=18-22℃ and H=60-70%, the optimization is to set ventilation to 70%RH and dehumidification to stop, with N=64 and RMSE=3.2%RH. Heating power is linked to the rate of humidity increase: v h When the RH / min is 0.8%, P = 100 + 200 × (0.8 / 1.5) ≈ 206 W, and when T exceeds 38℃, it drops back to 165 W.

[0051] Safety Protection: During a typhoon and heavy rain, one main sensor short-circuited (humidity difference 7%RH). The system automatically switched to a slave sensor and used Kalman filtering to fuse the data. k Deviation 1.5%RH; Heating device I h =1.3A (rated 1.1A, overload 18%), drops to 0.8A (50% power) after 5 seconds, and I after 10 seconds h When the condensate level (h) is restored to 1.0A, the drain pump starts, and the platform simultaneously sends a "Strengthen Inspection During Typhoon" reminder.

[0052] 2. Performance Comparison Data Table 2: Comparison of Anti-condensation Effects of Low-Voltage Switchgear in High-Temperature and High-Humidity Coastal Industrial Zones Typhoon humidity forecast accuracy ±9.5%RH ±3.5%RH Wind speed correction + η dynamic adjustment Summer high load energy consumption ratio 1:1.8 (Dehumidification: Heat dissipation) 1:1.2 Load linkage power optimization Average lifespan of electrical components 6.2 years 8.5 years Precise temperature and humidity control Typhoon season fault recovery time 45 minutes / session 8 minutes / session Fault self-diagnosis + emergency response strategy Annual energy consumption for preventing condensation 1280kWh / unit 760kWh / unit Coordinated operation of actuators Explanation: Table 2 is based on one year of operating data from 10 switchgear units in an industrial area. Traditional systems do not consider the impact of wind speed on ventilation efficiency during typhoons, with a prediction accuracy of ±9.5%RH. During high summer loads, dehumidification and heat dissipation conflict, resulting in an energy consumption ratio of 1:1.8, and the average lifespan of components due to temperature and humidity fluctuations is only 6.2 years. This invention, through dynamic correction using η, achieves an accuracy of ±3.5%, optimizes the energy consumption ratio to 1:1.2 based on load linkage, and extends component lifespan to 8.5 years. Traditional systems require 45 minutes to recover from typhoon season faults (manual troubleshooting), while this invention, with its self-diagnosis and emergency strategy, recovers in 8 minutes. Annual energy consumption is reduced from 1280kWh to 760kWh, meeting the continuous production needs of industrial areas for reliable power supply and energy conservation, while also mitigating the impact of extreme weather such as typhoons.

[0053] refer to Figure 2This figure visually demonstrates the advantages of this invention in humidity prediction at high altitudes. Traditional systems do not perform saturated humidity correction for changes in air pressure at high altitudes. As altitude increases, the calculation deviation for air saturated humidity increases, causing the 30-minute humidity prediction error to rise from ±8.5%RH at 0m in plains areas to ±12.0%RH at 2000m, severely impacting the advance notice and accuracy of anti-condensation strategies. This invention introduces an air pressure correction formula... The saturated humidity benchmark is dynamically adjusted to keep the prediction error at various altitudes stable within the range of ±3.2%-±4.0%RH. Even in a high-altitude scenario of 1800m (as in Example 1), the error is only ±3.8%RH, providing a reliable basis for accurately activating the anti-condensation actuator and solving the technical pain point of "inaccurate prediction at high altitudes" in traditional systems.

[0054] refer to Figure 3 The diagram clearly demonstrates the temperature control advantages of the "system linkage module" in this invention. During high-load summer periods (such as when the bus current approaches its rated value in Example 2), traditional systems, due to the combined heating of the heating device and electrical components and the lack of load linkage logic, cause the cabinet temperature to rise from 35°C to 47°C within 40 minutes, far exceeding the 40°C safety threshold. This can accelerate the aging of insulation materials and even lead to malfunctions. This invention, by acquiring load data from the power monitoring system, dynamically adjusts the heating power and prioritizes ventilation and heat dissipation when the component load exceeds 80% of its rated value. This keeps the cabinet temperature stable within the 35-37°C range, meeting both the temperature requirements for condensation prevention and ensuring that electrical components operate within a safe temperature range, thus solving the problem of "conflict between condensation prevention and component heat dissipation" in traditional systems.

[0055] refer to Figure 4 This figure highlights the energy-saving effect of the "parameter self-learning + actuator coordination" of this invention. Traditional systems lack seasonal adaptability and coordination logic in their actuators: in summer, they rely solely on high-power dehumidification (energy consumption of 480kWh), and in winter, they still use fixed-power heating, resulting in a total annual energy consumption of 1280kWh. This invention, through a parameter self-learning module, optimizes the power ratio of dehumidification and ventilation for high temperature and humidity in summer (such as 60% dehumidification power + dynamic fan speed regulation in Example 2), and reduces the starting frequency of the heating device in winter, reducing summer energy consumption to 360kWh and winter energy consumption to 170kWh, resulting in a total annual energy consumption reduction of 320kWh, while ensuring anti-condensation effect, achieving a balance between energy saving and reliability, and solving the defects of traditional systems that are "high in energy consumption and lack seasonal adaptability".

[0056] refer to Figure 5This figure verifies the reliability advantages of the "master-slave redundancy + Kalman filtering" approach in this invention. Traditional systems rely on a single sensor for data acquisition, and after a failure, the data deviation increases rapidly over time, reaching 18%RH after 20 seconds, far exceeding the reliable threshold of 3%RH, causing the anti-condensation strategy to become completely inaccurate. This invention employs a master-slave sensor redundancy arrangement. When the master sensor fails (e.g., the humidity difference exceeds 5%RH in Example 1), it automatically switches to the slave sensor and fuses the master and slave data using a Kalman filtering algorithm. This ensures that the data deviation remains ≤2%RH within 20 seconds after a failure, far below the reliable threshold, ensuring that the system can still accurately monitor the humidity inside the cabinet even when a sensor fails, avoiding the risk of "blind control" and filling the technical gap in traditional systems where "sensor failures lack redundancy backup."

[0057] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A low-voltage switchgear anti-condensation and humidity control system, characterized in that, include: The data acquisition module collects real-time humidity H, internal temperature T, external temperature T0, external humidity H0, and temperature gradient change rate k on the surface of the low-voltage switchgear. The temperature gradient change rate k is calculated by the ratio of the temperature difference between two adjacent temperature sensors to the installation distance. The humidity prediction module, based on the collected data, combined with the structural parameters L and ventilation porosity ρ of the switchgear, as well as historical operating data, uses a long short-term memory neural network algorithm to predict the humidity change trend inside the cabinet within the next 30 minutes and outputs a predicted humidity curve; when the humidity inside the cabinet is in the range of 40%RH-70%RH, a condensation risk pre-assessment is initiated and the results are generated. When predicting humidity change trends, the humidity prediction module also introduces a ventilation efficiency correction factor η, which is calculated as follows: η=(v×ρ×A) / (L×S), where v is the air velocity outside the cabinet, ρ is the ventilation porosity, A is the area of ​​a single ventilation hole, L is the effective ventilation length of the cabinet, and S is the ventilation side area of ​​the cabinet. When the outdoor wind force is level 3-4, the air velocity v outside the cabinet will fluctuate between 0.5-3m / s. At this time, the v value is collected in real time by a wind speed sensor and substituted into the calculation of η. Then, η is weighted and fused with the output value of the intermediate layer of the long short-term memory neural network. The weight coefficient α is obtained by training based on the correlation between historical wind speed and prediction error. The anti-condensation module selectively activates heating, ventilation, or dehumidification devices based on pre-assessment results; when the predicted humidity exceeds 85%RH and lasts for more than 10 minutes, the heating device is activated first. The system linkage module is used to interact with the power monitoring system of the low-voltage switchgear to obtain the operating load P of the electrical components inside the switchgear. l When the load on electrical components exceeds 80% of the rated load, the operating parameters of the anti-condensation actuator are adjusted in conjunction with the load. During peak load periods, if the heating device is running, the heating power is increased; if the ventilation device is running, the fan speed is increased. When the system linkage module interacts with the power monitoring system, it obtains the operating load P of the electrical components. l Surface temperature T e The system monitors the circuit breaker's open / closed status and the bus current Im. When the circuit breaker is closed and the bus current Im exceeds 70% of the rated current, if the anti-condensation actuator is operating the heating device, the system will adjust the upper limit of the heating power from 300W to 250W, and adjust the power according to T. e The numerical values ​​are dynamically adjusted. In addition, when the power monitoring system detects an overload warning for a certain section of the bus, the system linkage module will adjust the anti-condensation strategy in advance and reduce the humidity inside the cabinet to below 60%RH 5 minutes before the overload occurs. The parameter self-learning module iteratively optimizes the neural network model parameters of the humidity prediction module based on historical operating data and actual anti-condensation effects, while recording the optimal start-up threshold and operating parameters of each actuator under different seasons and environments. The safety protection module monitors the surface temperature T of the heating device in real time. h When T h When the temperature exceeds 70°C, the heating power is forcibly reduced to 50W and an early warning is issued; the water level h in the condensate collection box of the dehumidifier is monitored, and when h exceeds 80% of the total capacity, the drain pump is started and maintenance is prompted; at the same time, in the event of a communication failure or abnormality of the actuator, the system automatically switches to the preset emergency anti-condensation strategy.

2. The anti-condensation and humidity control system for low-voltage switchgear according to claim 1, characterized in that, It also includes a condensation risk quantification assessment unit, which determines the likelihood of condensation occurring by calculating a condensation risk coefficient C, using the formula: C=(HH s )×(T0-T) / (T0+273.15), where H is the real-time humidity inside the cabinet. s The saturation humidity is the humidity at the current cabinet temperature T, where T0 is the outside temperature and T is the inside temperature. When C is greater than 0.8, it is considered a high risk of condensation. The system will immediately start the heating device and increase the power to 250W, while simultaneously starting the ventilation device at a speed of 2000r / min. The system will continuously monitor the humidity changes inside the cabinet and update the condensation risk coefficient every 5 seconds. If C is between 0.5 and 0.8, it is considered a medium risk of condensation. The system will prioritize starting the dehumidification device and calculate the C value every 10 seconds. If C is less than 0.5, it is judged as a low risk of condensation, and only the ventilation device is activated for intermittent ventilation at a speed of 1500 r / min.

3. The anti-condensation and humidity control system for low-voltage switchgear according to claim 1, characterized in that, The data acquisition module also collects the air pressure value inside the switchgear. The humidity prediction module combines air pressure values ​​when predicting humidity. For saturated humidity Make corrections, the corrected version ,in Standard atmospheric pressure This refers to the real-time air pressure inside the cabinet; if the real-time air pressure inside the cabinet... At a pressure of 84 kPa and a current cabinet temperature T of 25°C, the uncorrected saturated humidity is... It is 23.04 g / m³, the corrected =23.04×(101.325 / 84)≈23.04×1.206≈27.8g / m³.

4. The anti-condensation and humidity control system for low-voltage switchgear according to claim 1, characterized in that, The parameter self-learning module employs a hierarchical learning strategy when optimizing the long short-term memory neural network model of the humidity prediction module: first, historical data is categorized by season; then, for each season's data, the number of hidden layer neurons N and the learning rate η of the neural network are optimized separately. l The root mean square error (RMSE) is used as the evaluation index for model accuracy, and the calculation method is as follows: H ip Let i be the humidity value predicted in the i-th time. Let be the actual humidity value of the i-th time, and n be the number of samples; when the RMSE of a certain season is greater than 5%RH, the model for that season is re-optimized.

5. The anti-condensation and humidity control system for low-voltage switchgear according to claim 1, characterized in that, The safety protection module monitors the surface temperature T of the heating device. h 1. Water level h in the condensate collection box of the dehumidifier; 2. Operating current I of the heating device. h Voltage U h Ventilation device fan speed n x Dehumidifier operating power P c， When heating device I h With rated current I hn If the ratio exceeds 1.2 and persists for 5 seconds, it is determined to be an overload fault. The system first reduces the heating power to 50% of the rated power and monitors I. h 10 seconds later I h If the current still exceeds 1.1 times the rated current, the power supply will be cut off, and emergency ventilation and dehumidification strategies will be activated; when the ventilation device n x With the set speed n xs If the deviation exceeds ±20% and lasts for 10 seconds, it is determined to be a fan failure. The system should switch to the backup ventilation duct or increase the dehumidifier power to 80%. If the dehumidifier P... c With rated power P cn If the ratio is less than 0.3 and the humidity inside the cabinet increases, it is determined that the dehumidification device has failed, and the system immediately starts the heating and ventilation device coordinated mode; the safety protection module divides the fault type into three levels: warning, fault, and emergency fault, which correspond to different audible and visual alarm signals and remote notification strategies.

6. The anti-condensation and humidity control system for low-voltage switchgear according to claim 1, characterized in that, The dehumidifier control in the anti-condensation module employs a multi-parameter coordinated adjustment strategy based on humidity, temperature, and air pressure: dehumidification power P c The adjustment formula is P c =P cn ×[0.6+0.2×(H / 100)+0.1×(T / 30)+0.1×(P a0 / P a )], where P cn H is the rated power of the dehumidifier, T is the real-time humidity inside the cabinet, and P is the temperature inside the cabinet. a P represents the real-time air pressure inside the cabinet. a0 The standard atmospheric pressure is used. The real-time calculation of the condensate discharge rate V is based on the formula V=0.05×t×(H1-40), where t is the dehumidification time and H1 is the humidity inside the cabinet when dehumidification is started. The coordinated switching logic between the dehumidification device and the heating and ventilation devices is as follows: when the temperature inside the cabinet T is below 15℃, the heating device is started first instead of the dehumidification device to avoid frost formation inside the cabinet due to dehumidification at low temperatures. When the ventilation device is running, if the humidity outside the cabinet H0 is lower than the real-time humidity inside the cabinet H and the difference between the temperature outside the cabinet T0 and the temperature inside the cabinet T is less than 5℃, the running time of the ventilation device is increased.

7. The anti-condensation and humidity control system for low-voltage switchgear according to claim 1, characterized in that, The temperature and humidity sensors in the data acquisition module adopt a redundant "master-slave" arrangement. Each monitoring area is equipped with one master sensor and at least one slave sensor. When the difference between the measured values ​​of the master and slave sensors exceeds 5%RH or 2℃, the sensor fault diagnosis program is automatically activated. This program uses a Kalman filter algorithm to fuse the data from the master and slave sensors and calculate the fused data. ,in For the fused data at time k, For the fused data at time k-1, For Kalman gain, This is the current sensor measurement value. The observation matrix is ​​used; if the deviation between the fused data and the main sensor data still exceeds 3%RH or 1℃, the main sensor is determined to be faulty, and the system automatically switches to the slave sensor for data acquisition. The main sensor is then marked as faulty, and a sensor fault report is generated through the edge computing unit and pushed to the operation and maintenance platform.

Images