Power distribution station indoor environment monitoring method and management system based on internet of things

Abstract

The application discloses a power distribution station indoor environment monitoring method and treatment system based on the Internet of Things, relates to the power distribution station environment treatment technical field, and comprises the following steps: collecting indoor temperature and humidity, equipment load current and outdoor temperature and humidity of each subzone of the power distribution station, and calculating the absolute humidity difference value between the inside and outside of each subzone; inputting the collected data and the difference value into a time sequence prediction network, and outputting a future time period temperature and humidity prediction sequence; distributing the air inlet and air outlet roles of the wall fan according to the heat load distribution, identifying the reverse time according to the target difference value sign and the prediction sequence; and establishing a four-state control decision with the target difference value sign and the reverse time as dimensions. The application realizes the dynamic distribution of the fan roles according to the heat load, the advance adjustment strategy according to the humidity reverse prediction, and the maintenance of the micro-positive pressure to inhibit the gap penetration condensation, and is convenient and reliable.

Description

Technical Field

[0001] This invention relates to the field of environmental management technology for power distribution stations, specifically to an indoor environmental monitoring and management system for power distribution stations based on the Internet of Things. Background Technology

[0002] Substations are critical nodes in power transmission and distribution, and their indoor environment directly affects the safe operation and lifespan of electrical equipment. Transformers, switchgear, and other equipment within substations continuously generate heat and are affected by changes in outdoor temperature and humidity, easily leading to condensation on walls and cabinet surfaces. This can cause insulation degradation or even short circuits. Therefore, real-time monitoring and effective management of the substation's indoor environment are crucial for ensuring power supply reliability.

[0003] In existing technologies, environmental management in power distribution stations typically uses IoT sensors to collect indoor temperature and humidity data. When humidity exceeds a preset threshold, wall-mounted fans are activated for ventilation or dehumidifiers are activated. In recent years, some solutions have introduced time-series predictive networks to forecast temperature and humidity trends, activating fans in advance when humidity is predicted to exceed the limit, thus providing a degree of foresight compared to traditional threshold alarm methods.

[0004] However, the above solution has the following shortcomings: First, the fan only acts as a passive terminal after the humidity exceeds the limit, and its operation direction of forward air intake or reverse air exhaust is fixed. It fails to dynamically allocate the role of the fan according to the difference in heat load of equipment in different zones of the substation, resulting in hot air spreading between zones and failing to be discharged in a directional manner. Second, the control strategy only relies on the comparison between the current humidity and the threshold, which is a passive response mode of "starting when the threshold is exceeded". When the outdoor humidity is higher than the indoor humidity, starting the fan will introduce humid air and aggravate condensation. It also cannot identify the time window when the absolute humidity difference between the inside and outside of each zone is about to reverse so as to adjust the strategy in advance. Third, it does not maintain a slight positive pressure in the room by controlling the difference in frequency between the intake and exhaust sides. There is no directional air pressure barrier at the wall gaps, and outdoor humid air can seep in without organization from the gaps to form local condensation dead zones. Summary of the Invention

[0005] To address the above problems, this invention provides a method and system for monitoring and managing the indoor environment of power distribution stations based on the Internet of Things (IoT).

[0006] The technical solution adopted in this invention is: an Internet of Things-based method for monitoring the indoor environment of a power distribution station, comprising:

[0007] S1. Collect data on indoor temperature and humidity of each zone of the substation, outdoor temperature and humidity of the corresponding wall outside each zone, and equipment load current. Calculate the absolute humidity difference between the inside and outside of each zone and take the minimum value of the absolute humidity difference between the inside and outside of each zone as the target absolute humidity difference between the indoor and outdoor areas of the substation.

[0008] S2. Input the collected data and absolute humidity difference into the time series prediction network, and output the temperature and humidity prediction sequence of each zone for future time periods.

[0009] S3. Determine the heat load distribution and sort the heat loads according to the load current of each zone's equipment. Assign the side wall fans of the zones in the first half of the heat load sorting sequence as exhaust fans and the side wall fans of the zones in the second half of the heat load sorting sequence as intake fans, forming an exhaust side wall fan set and an intake side wall fan set. At the same time, identify the reversal time based on the sign of the target absolute humidity difference and in combination with the temperature and humidity prediction sequence. The reversal time is the moment when the sign of the target difference changes from positive to negative.

[0010] S4. When the target difference is positive and the reversal moment has not been identified, each wall fan is operated according to the assigned role to make the sum of the frequency converters on the air intake side higher than the sum of the frequency converters on the air exhaust side in order to maintain a slight positive pressure in the substation room.

[0011] When the moment when the target difference is about to reverse from positive to negative is identified, the sum of the frequency of the exhaust side inverter is gradually increased and the sum of the frequency of the intake side inverter is simultaneously decreased to accelerate dehumidification. The sum of the frequency of the intake side inverter is kept not lower than the sum of the frequency of the exhaust side inverter to maintain a slight positive pressure in the substation room. Before the reversal moment is reached, all wall fans are shut down and the dehumidifier is activated.

[0012] S5. When the target difference is negative or zero and the reversal moment is not detected, shut down all wall fans and activate the dehumidifier.

[0013] When the moment when the target difference is about to turn from negative to positive is detected, the system remains closed until the moment of reversal is reached. Each wall fan is operated according to its assigned role to ensure that the sum of the frequency converters on the air intake side is higher than the sum of the frequency converters on the air exhaust side in order to maintain a slight positive pressure in the substation room.

[0014] Preferably, the step of identifying the reversal time based on the sign of the target absolute humidity difference and in conjunction with the temperature and humidity prediction sequence includes:

[0015] After detecting a change in the sign of the target absolute humidity difference in the temperature and humidity prediction sequence, the target difference sign in subsequent prediction time steps is continuously monitored. Only when the changed target difference sign is continuously maintained for more than the preset confirmation time is it determined to be a valid reversal moment.

[0016] The countdown is determined based on the time distance between the effective reversal time and the current time. When the countdown is lower than the preset warning threshold, the corresponding control strategy is triggered.

[0017] Preferably, it also includes role adjustment A and role adjustment B.

[0018] The role adjustment A is as follows: the heat load sorting is updated according to a preset inference cycle. When the updated heat load sorting is consistent with the sorting before the change within multiple consecutive preset inference cycles and the cumulative duration exceeds the preset stable duration, the role allocation is switched according to the updated heat load sorting. If the preset stable duration is not exceeded, the original role remains unchanged.

[0019] The role adjustment B is as follows: when there are at least two zones with heat load differences lower than the preset heat load balance threshold, the roles are reassigned according to the cumulative running time of the corresponding side wall fans, and the wall fans with shorter cumulative running time are assigned as air intake roles.

[0020] Role adjustment A is used for role adjustment when the heat load sequence changes, and role adjustment B is used for role adjustment when the heat loads of adjacent zones are close. When both role adjustment A and role adjustment B are satisfied, role adjustment B is executed first.

[0021] The heat load difference is the absolute difference between the heat load characterization values ​​of the two zones, which are determined by the zone equipment load current or the product of the square of the zone equipment load current and the equivalent resistance.

[0022] Preferably, the data acquisition further includes the indoor-outdoor pressure difference, and maintaining a slight positive pressure indoors at the substation includes:

[0023] The frequency converters on the air inlet and exhaust sides are adjusted in a closed loop based on the indoor and outdoor pressure difference. The sum of the frequency converters on the air inlet side is the sum of the frequency converters of all wall fans in the air inlet side wall fan assembly, and the sum of the frequency converters on the exhaust side is the sum of the frequency converters of all wall fans in the exhaust side wall fan assembly. The indoor and outdoor pressure difference is maintained between 5 and 15 Pa to ensure that the airflow flows along a directional path from the air inlet side to the exhaust side and to suppress local condensation caused by unorganized infiltration from wall gaps. When the pressure difference is less than 5 Pa, the sum of the frequency converters on the air inlet side is increased and the sum of the frequency converters on the exhaust side is decreased. When the pressure difference is greater than 15 Pa, the sum of the frequency converters on the air inlet side is decreased and the sum of the frequency converters on the exhaust side is increased.

[0024] The maintenance of a slightly positive pressure in the substation also includes a slightly positive pressure constraint during the accelerated dehumidification phase: when the moment when the target difference is about to reverse from positive to negative is detected, the accelerated dehumidification phase is entered; during the accelerated dehumidification phase, the frequency of the variable frequency drive on the air intake side and the frequency of the variable frequency drive on the air exhaust side are synchronously adjusted to keep the sum of the variable frequency drives on the air intake side not lower than the sum of the variable frequency drives on the air exhaust side as a constraint; when the difference in the sum of the variable frequency drives shrinks to 0Hz, the accelerated dehumidification is stopped, all wall fans are turned off, and the dehumidifier is activated.

[0025] Preferably, the step of gradually increasing the frequency of the exhaust side inverter to accelerate dehumidification includes:

[0026] The heat generation of the equipment is calculated based on the load current of each zone's equipment. The baseline ventilation volume required to maintain a stable indoor temperature in the substation is then calculated based on the current equipment heat generation and the outdoor temperature. The baseline frequency conversion of all wall fans is determined based on the baseline ventilation volume. During the accelerated dehumidification phase, the acceleration ratio is determined by combining the countdown time and the difference between the current target absolute humidity. The shorter the countdown and the smaller the difference, the larger the acceleration ratio. The acceleration ratio is constrained to have an upper limit of 0Hz, with the difference between the current sum of the frequency conversions on the intake side and the current sum of the frequency conversions on the exhaust side not being less than 0Hz. Based on the acceleration ratio, the frequency conversion on the exhaust side is increased and the frequency conversion on the intake side is decreased simultaneously on the basis of the baseline frequency conversion, keeping the sum of the frequency conversions on the intake side not lower than the sum of the frequency conversions on the exhaust side to maintain a slight positive pressure in the substation.

[0027] Preferably, the IoT-based indoor environmental management system for power distribution stations includes:

[0028] The multi-source sensing module is used to collect data on indoor temperature and humidity of each zone of the substation, outdoor temperature and humidity of the corresponding wall outside each zone, and equipment load current. It calculates the absolute humidity difference between the inside and outside of each zone and takes the minimum value of the absolute humidity difference between the inside and outside of each zone as the target absolute humidity difference between the indoor and outdoor areas of the substation.

[0029] The trend prediction module is used to input the collected data and absolute humidity difference into the time series prediction network and output the temperature and humidity prediction sequence of each zone for future time periods.

[0030] The airflow management decision module is used to determine the heat load distribution based on the load current of each zone's equipment and to allocate the air intake and exhaust roles of the wall fans. It identifies the reversal time based on the target absolute humidity difference sign and the temperature and humidity prediction sequence, and generates management instructions for the wall fans based on the target difference sign and the reversal time.

[0031] The wall-mounted fan control module is used to control the forward and reverse rotation and frequency conversion of multiple wall-mounted fans on the wall of the substation according to the control instructions, so that the sum of the frequency conversion on the air intake side is higher than the sum of the frequency conversion on the exhaust side to maintain a slight positive pressure in the substation.

[0032] Preferably, the multi-source sensing module deploys temperature and humidity sensors in pairs on both the inner and outer sides of the wall of each partition, so that the sign of the absolute humidity difference reflects the direction of humid air penetration through the wall on that side; the equipment load current is collected from the feeder circuit of each partition; the collected data is filtered by sliding window mean and then output to the trend prediction module.

[0033] Preferably, the trend prediction module executes the time-series prediction network with a preset inference cycle, and outputs an hourly temperature and humidity prediction sequence covering a preset future time period each time. The prediction sequence includes hourly predicted values ​​of the absolute humidity difference, so that the airflow governance decision module can detect changes in the sign of the target difference step by step.

[0034] Preferably, the airflow management decision module establishes a four-state control state machine based on the identification results of the target difference sign and the reversal time. The four-state control state machine includes: a ventilation and dehumidification state when the target difference is positive and the reversal time has not been identified; an accelerated dehumidification state when the target difference is positive and the reversal time from positive to negative has been identified; a closed dehumidification state when the target difference is negative or zero and the reversal time has not been identified; and a waiting ventilation state when the target difference is negative or zero and the reversal time from negative to positive has been identified.

[0035] The airflow management decision module generates corresponding management instructions based on the current state. The management instructions include the role identification, operating direction and frequency of each wall fan. The airflow management decision module also receives the cumulative running time of each wall fan from the wall fan management module. When there are at least two zones with heat load differences lower than the preset heat load balancing threshold, a role reassignment instruction is generated based on the cumulative running time.

[0036] The four-state control state machine performs a state transition when the identification result of the target difference sign or reversal moment changes, and updates the role identifier, running direction and frequency conversion frequency in the governance instruction when the role reassignment instruction changes.

[0037] Preferably, the wall-mounted ventilation fan control module includes:

[0038] The frequency control unit outputs frequency commands to the frequency converter based on the frequency parameters in the management instructions;

[0039] The forward / reverse switching unit switches between forward air intake and reverse air exhaust according to the role identifier in the management instruction and performs electrical interlock protection during the switching process;

[0040] The role adjustment unit records the cumulative runtime of each wall fan and its current role, and feeds it back to the airflow management decision module for role reassignment.

[0041] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0042] This invention proposes a method to dynamically allocate the roles of wall-mounted fans for air intake and exhaust based on the heat load of each zone's equipment. This allows hot air from zones with high heat loads to be directly exhausted outdoors via the corresponding exhaust fan, suppressing the diffusion of hot air between zones and replacing disordered ventilation with directional heat removal. Furthermore, it proposes identifying reversal moments based on the sign of the target absolute humidity difference and the temperature and humidity prediction sequence. A four-state control mechanism determines corresponding treatment strategies for different combinations of target difference signs and reversal moments, enabling accelerated dehumidification when the absolute humidity difference between the inside and outside of each zone is about to turn from positive to negative, and... Before the rotation, the fan is shut off and the dehumidifier is activated, preventing outdoor humid air from entering the room and exacerbating condensation. When the target difference is about to turn from negative to positive, the fan remains off until ventilation is activated after the rotation, thus realizing the proactive use of the humidity reversal time window and replacing the passive threshold response mode. The method of making the frequency converter on the air intake side higher than that on the exhaust side to maintain a slight positive pressure in the substation room is proposed, which realizes the directional flow of air along the air intake side to the exhaust side, suppresses the local condensation dead zone caused by the unorganized infiltration of outdoor humid air from wall gaps, and makes the controllable air pressure barrier replace the protection without pressure difference, which is reliable. Attached Figure Description

[0043] Figure 1 This is a flowchart of the method of the present invention;

[0044] Figure 2 This is the four-state control decision logic diagram of the present invention;

[0045] Figure 3 This is a system framework diagram of the present invention. Detailed Implementation

[0046] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0047] Reference Figure 1 As shown, the IoT-based indoor environment monitoring method for power distribution stations includes:

[0048] S1. Collect data on indoor temperature and humidity of each zone of the substation, outdoor temperature and humidity of the corresponding wall outside each zone, and equipment load current. Calculate the absolute humidity difference between the inside and outside of each zone and take the minimum value of the absolute humidity difference between the inside and outside of each zone as the target absolute humidity difference between the indoor and outdoor areas of the substation.

[0049] Regarding the calculation of the absolute humidity difference between the inside and outside of each zone, the first step is to use the temperature data collected by the indoor temperature and humidity sensor. and relative humidity and the temperature collected by the outdoor temperature and humidity sensor. and relative humidity Calculate the absolute humidity for both indoor and outdoor environments. The calculation of absolute humidity is based on the Magnus saturated vapor pressure formula, which is as follows:

[0050]

[0051] Where 6.112 is the reference value of saturated vapor pressure at 0°C and standard atmospheric pressure (in hPa), 17.67 and 243.5 are Magnus formula constants, 2.1674 is the conversion factor from vapor pressure to absolute humidity, 273.15 is the offset from the Celsius temperature scale to the Kelvin temperature scale, T is the Celsius temperature, RH is the relative humidity percentage, and AH is the absolute humidity, in g / m³. The absolute humidity difference between the inside and outside of each zone is also shown. ,when A positive value indicates that the indoor absolute humidity is higher than the outdoor humidity, and ventilation is beneficial for dehumidification; when... A value of negative or zero indicates that the outdoor absolute humidity is not lower than the indoor humidity, and starting ventilation will introduce humid air.

[0052] S2. Input the collected data and absolute humidity difference into the time series prediction network, and output the temperature and humidity prediction sequence of each zone for future time periods.

[0053] The time-series prediction network employs a two-layer LSTM network, with 64 hidden units in each layer. The network input includes indoor temperature and humidity data from the past hour, equipment load current, absolute humidity difference, and current outdoor temperature and humidity. The output is a time-step prediction sequence of temperature and humidity for each zone over the next two hours, with a prediction time step of 5 minutes and a total of 24 prediction time steps. The prediction sequence includes time-step predictions of the absolute humidity difference to allow subsequent steps to detect changes in the sign of the target difference.

[0054] S3. Determine the heat load distribution and sort the heat loads according to the load current of each zone's equipment. Assign the side wall fans of the zones in the first half of the heat load sorting sequence as exhaust fans and the side wall fans of the zones in the second half of the heat load sorting sequence as intake fans, forming an exhaust side wall fan set and an intake side wall fan set. At the same time, identify the reversal time based on the sign of the target absolute humidity difference and in combination with the temperature and humidity prediction sequence. The reversal time is the moment when the sign of the target difference changes from positive to negative.

[0055] The heat load distribution is determined based on the load current of each zone's equipment. The wall fans corresponding to zones in the first half of the heat load sequence are assigned as exhaust fans, and those in the second half are assigned as intake fans. Since the heat output of the electrical equipment is proportional to the square of the load current, the heat load sequence can be directly determined by the magnitude of the load current in each zone. Assigning the wall fans corresponding to zones in the first half of the heat load sequence as exhaust fans allows hot air from those zones to be directly discharged outdoors, inhibiting diffusion to other zones. Assigning the wall fans corresponding to zones in the second half of the heat load sequence as intake fans allows drier outdoor air to enter the room through the intake side, forming a directional airflow. The exhaust fans rotate in reverse for exhaust, and the intake fans rotate in forward for intake.

[0056] When the number of partitions is even, the first and second halves of the sequence are divided equally according to heat load. When the number of partitions is odd, the partitions in the middle of the sequence are assigned to the second half, making the second half have one more partition than the first half. For example, when there are 5 partitions, the partitions ranked 1st and 2nd by heat load constitute the first half, and the partitions ranked 3rd to 5th by heat load constitute the second half.

[0057] Reference Figure 2 As shown, the sign of the absolute humidity difference is detected step-by-step in the temperature and humidity prediction sequence. When the sign changes from positive to negative or from negative to positive, the corresponding time step is recorded as a candidate reversal time. Reversal times include two types: a sign about to turn from positive to negative indicates the dehumidification window is about to close, and a sign about to turn from negative to positive indicates the ventilation and dehumidification window is about to open. The specific actions involve four stages: dehumidification and ventilation, accelerated dehumidification, dehumidification maintenance, and transitional start-up.

[0058] S4. When the target difference is positive and the reversal moment has not been identified, each wall fan is operated according to the assigned role to make the sum of the frequency converters on the air intake side higher than the sum of the frequency converters on the air exhaust side in order to maintain a slight positive pressure in the substation room.

[0059] When the moment when the target difference is about to reverse from positive to negative is identified, the sum of the frequency of the exhaust side inverter is gradually increased and the sum of the frequency of the intake side inverter is simultaneously decreased to accelerate dehumidification. The sum of the frequency of the intake side inverter is kept not lower than the sum of the frequency of the exhaust side inverter to maintain a slight positive pressure in the substation room. Before the reversal moment is reached, all wall fans are shut down and the dehumidifier is activated.

[0060] When the target difference is positive and no reversal moment is detected, each wall fan operates according to its assigned role, ensuring the frequency of the intake side is higher than that of the exhaust side to maintain a slight positive pressure inside the substation. At this time, the indoor absolute humidity is higher than that outdoors, and the outdoor air is drier, making ventilation beneficial for dehumidification. The wall fans in the intake role are controlled to rotate forward for air intake, and the wall fans in the exhaust role are controlled to rotate in reverse for air exhaust, ensuring the frequency of the intake side is higher than that of the exhaust side to maintain a slight positive pressure inside the substation.

[0061] When the system detects the reversal point where the target difference is about to turn from positive to negative, it gradually increases the sum of the frequency converters on the exhaust side and simultaneously decreases the sum of the frequency converters on the intake side to accelerate dehumidification. It maintains that the sum of the frequency converters on the intake side is not lower than the sum of the frequency converters on the exhaust side to maintain a slight positive pressure in the substation. Before the reversal point is reached, all wall fans are shut down and the dehumidifier is activated. At this time, the outdoor humidity is about to exceed the indoor humidity, and the dehumidification window is about to close. The system synchronously adjusts the frequency converters on the intake and exhaust sides—gradually decreasing the frequency converter on the intake side and gradually increasing the frequency converter on the exhaust side, gradually reducing the difference in the sum of the frequency converters while maintaining that the intake side is not lower than the exhaust side, to maximize ventilation and dehumidification capacity under slight positive pressure constraints. When the difference in the sum of the frequency converters decreases to 0Hz, accelerated dehumidification stops, and all fans are shut down before the reversal point is reached to prevent outdoor humid air from entering. Simultaneously, the dehumidifier is activated to continue dehumidification in a sealed manner.

[0062] S5. When the target difference is negative or zero and the reversal moment is not detected, shut down all wall fans and activate the dehumidifier.

[0063] When the reversal point where the target difference is about to turn from negative to positive is detected, remain in the off state until the reversal point is reached. Operate the wall fans according to their assigned roles, ensuring that the sum of the inverter frequencies on the intake side is higher than the sum of the inverter frequencies on the exhaust side to maintain a slight positive pressure in the substation room. At this time, the outdoor humidity will soon be lower than the indoor humidity, but ventilation is still not advisable. The system remains in the off state until the reversal point is reached, at which point ventilation and dehumidification will be activated while maintaining a slight positive pressure.

[0064] By adopting the above technical solution, a four-state control strategy is formed by combining the sign of the target absolute humidity difference value with the reversal time. Each state corresponds to a deterministic treatment scheme consistent with the dehumidification target, and there is no control blind spot. In particular, when the dehumidification window is about to close, the dehumidification is accelerated to make full use of the remaining ventilation time, realizing the forward utilization of the humidity reversal time window and replacing the passive threshold response mode.

[0065] In the specific implementation process, taking a 10kV substation as an example, the substation is divided into two zones: the transformer area on the east side and the switchgear area on the west side. To collect data on indoor temperature and humidity, equipment load current, and outdoor temperature and humidity in each zone, temperature and humidity sensors are deployed on both the inner and outer sides of the walls of each zone. The sensors collect data every 5 minutes. The indoor temperature and humidity sensors are installed 1.5 meters above the ground on the inner side of the wall of each zone, and the outdoor temperature and humidity sensors are installed 1.5 meters above the ground on the outer side of the corresponding wall, ensuring that the inner and outer sensors are at the same horizontal height to accurately reflect the direction of humid air penetration through the wall. Equipment load current is collected from the feeder circuits of each zone, using current transformers with an accuracy of 0.5 class.

[0066] The step of identifying the reversal time based on the sign of the target absolute humidity difference and in combination with the temperature and humidity prediction sequence includes:

[0067] After detecting a change in the sign of the target absolute humidity difference in the temperature and humidity prediction sequence, the target difference sign in subsequent prediction time steps is continuously monitored. Only when the changed target difference sign is continuously maintained for more than the preset confirmation time is it determined to be a valid reversal moment.

[0068] The countdown is determined based on the time distance between the effective reversal time and the current time. When the countdown is lower than the preset warning threshold, the corresponding control strategy is triggered.

[0069] In practical implementation, a preset confirmation time of 10 minutes is used. This means that after the target difference sign changes, the target difference sign must maintain the same direction for two consecutive prediction time steps (each time step is 5 minutes) before it is considered a valid reversal moment. This technical solution eliminates false reversals caused by instantaneous fluctuations or noise in the prediction sequence, ensuring the reliability of reversal moment determination and avoiding unnecessary operating condition switching due to misjudgments. A preset warning threshold of 15 minutes is used. This means that when the countdown is less than 15 minutes, the corresponding control strategy (accelerated dehumidification or waiting for startup) is triggered, ensuring the system has sufficient response time to complete the operating condition switch.

[0070] The exhaust role corresponds to the wall fan reversing to exhaust air, and the air intake role corresponds to the wall fan rotating forward to intake air, so that the hot air in the zone with the heat load in the first half of the sequence is directly discharged to the outside through the corresponding side exhaust to suppress diffusion to other zones.

[0071] This also includes character adjustment A and character adjustment B.

[0072] The role adjustment A is as follows: the heat load sorting is updated according to a preset inference cycle. When the updated heat load sorting is consistent with the sorting before the change within multiple consecutive preset inference cycles and the cumulative duration exceeds the preset stable duration, the role allocation is switched according to the updated heat load sorting. If the preset stable duration is not exceeded, the original role remains unchanged.

[0073] The role adjustment B is as follows: when there are at least two zones with heat load differences lower than the preset heat load balance threshold, the roles are reassigned according to the cumulative running time of the corresponding side wall fans, and the wall fans with shorter cumulative running time are assigned as air intake roles.

[0074] Role adjustment A is used for role adjustment when the heat load sequence changes, and role adjustment B is used for role adjustment when the heat loads of adjacent zones are close. When both role adjustment A and role adjustment B are satisfied, role adjustment B is executed first.

[0075] The heat load difference is the absolute difference between the heat load characterization values ​​of the two zones, which are determined by the zone equipment load current or the product of the square of the zone equipment load current and the equivalent resistance.

[0076] The heat load difference here is represented by the absolute difference of the load current between the two zones; when the difference of the load current between at least two zones is less than 5A, it indicates that the heat load difference is not significant.

[0077] In the specific implementation process, the preset stability period is 10 minutes. The heat load ranking is dynamically updated according to a preset inference cycle, which can be consistent with the inference cycle of the time series prediction network, for example, 5 minutes. Within each preset inference cycle, the system re-acquires the load current of each zone's equipment and uses the zone's equipment load current or the product of the square of the zone's equipment load current and its equivalent resistance as the heat load characterization value, ranking each zone from high to low. The system compares the heat load ranking of the current cycle with that of the previous cycle; if the ranking change remains unchanged for more than 10 minutes, the system reassigns the intake and exhaust roles of the wall fans according to the current cycle's heat load ranking; if the ranking change does not remain unchanged for more than 10 minutes, the change is considered a short-term load fluctuation, and the original roles are maintained. When the heat load ranking changes, the system does not immediately switch role assignments but continuously monitors, only switching occurs if the change lasts for more than 10 minutes; otherwise, the original roles are maintained. This technical solution prevents frequent role switching caused by small fluctuations in load current and avoids the impact of frequent fan start-stop cycles on motor lifespan. A preset heat load balancing threshold of 5A is set. When the load current difference between at least two zones is less than 5A, it indicates that the heat load difference is not significant. At this time, the roles are reassigned based on the cumulative running time of the wall fans on the corresponding side, and the wall fans with the shorter cumulative running time are assigned to the air intake role. With the above technical solution, the air intake side draws in outdoor dust and experiences greater wear, while the exhaust side discharges hot air at a higher temperature. The operating loads of the two are different. By balancing and rotating the cumulative running time, the running time of each fan tends to be balanced, thus extending the overall fan life.

[0078] The data acquisition also includes the indoor-outdoor pressure difference, and maintaining a slight positive pressure indoors at the substation includes:

[0079] The frequency converters on the air inlet and exhaust sides are adjusted in a closed loop based on the indoor and outdoor pressure difference. The sum of the frequency converters on the air inlet side is the sum of the frequency converters of all wall fans in the air inlet side wall fan assembly, and the sum of the frequency converters on the exhaust side is the sum of the frequency converters of all wall fans in the exhaust side wall fan assembly. The indoor and outdoor pressure difference is maintained between 5 and 15 Pa to ensure that the airflow flows along a directional path from the air inlet side to the exhaust side and to suppress local condensation caused by unorganized infiltration from wall gaps. When the pressure difference is less than 5 Pa, the sum of the frequency converters on the air inlet side is increased and the sum of the frequency converters on the exhaust side is decreased. When the pressure difference is greater than 15 Pa, the sum of the frequency converters on the air inlet side is decreased and the sum of the frequency converters on the exhaust side is increased.

[0080] The maintenance of a slightly positive pressure in the substation also includes a slightly positive pressure constraint during the accelerated dehumidification phase: when the moment when the target difference is about to reverse from positive to negative is detected, the accelerated dehumidification phase is entered; during the accelerated dehumidification phase, the frequency of the variable frequency drive on the air intake side and the frequency of the variable frequency drive on the air exhaust side are synchronously adjusted to keep the sum of the variable frequency drives on the air intake side not lower than the sum of the variable frequency drives on the air exhaust side as a constraint; when the difference in the sum of the variable frequency drives shrinks to 0Hz, the accelerated dehumidification is stopped, all wall fans are turned off, and the dehumidifier is activated.

[0081] In the specific implementation process, a micro-differential pressure sensor is used to collect the indoor and outdoor pressure difference with an accuracy of 0.1 Pa. When the pressure difference is below 5 Pa, the frequency of the inverter on the air inlet side is increased by 5 Hz and the frequency of the inverter on the air outlet side is decreased by 3 Hz. When the pressure difference is above 15 Pa, the frequency of the inverter on the air inlet side is decreased by 5 Hz and the frequency of the inverter on the air outlet side is increased by 3 Hz. After adjustment, the frequency of the inverter on the air inlet side is always higher than that on the air outlet side, maintaining a frequency difference of at least 3 Hz. Using the above technical solution, the micro-positive pressure causes the airflow to flow along a directional path from the air inlet side to the air outlet side, suppressing local condensation caused by unorganized infiltration from wall gaps. When the indoor and outdoor pressure difference is 0 Pa, outdoor humid air can freely infiltrate from wall gaps, forming a local high-humidity area near the gaps, leading to condensation. After maintaining a micro-positive pressure of 5 to 15 Pa, the airflow at the gaps flows from the inside to the outside, preventing the unorganized infiltration of outdoor humid air.

[0082] The stepwise increase of the frequency converter frequency on the exhaust side to accelerate dehumidification includes:

[0083] The heat generation of the equipment is calculated based on the load current of each zone's equipment. The baseline ventilation volume required to maintain a stable indoor temperature in the substation is then calculated based on the current equipment heat generation and the outdoor temperature. The baseline frequency conversion of all wall fans is determined based on the baseline ventilation volume. During the accelerated dehumidification phase, the acceleration ratio is determined by combining the countdown time and the difference between the current target absolute humidity. The shorter the countdown and the smaller the difference, the larger the acceleration ratio. The acceleration ratio is set at the upper limit of the constraint that the difference between the current sum of the frequency conversions on the air intake side and the current sum of the frequency conversions on the air exhaust side should not be less than 0Hz. Based on the acceleration ratio, the frequency conversion of the air exhaust side is increased simultaneously and the frequency conversion of the air intake side is decreased simultaneously on the basis of the baseline frequency conversion, keeping the sum of the frequency conversions on the air intake side not lower than the sum of the frequency conversions on the air exhaust side to maintain a slight positive pressure in the substation.

[0084] In the specific implementation process, the current method for calculating the heat generation of the equipment is as follows:

[0085]

[0086] Where Q represents the current heat output of the equipment in the substation. For the load current of each zone, This is the equivalent resistance of the corresponding circuit.

[0087] The calculation method for the baseline ventilation volume is as follows:

[0088]

[0089] Where V is the baseline ventilation volume. Take the air density as 1.2 kg / m³. Take the specific heat capacity of air as 1005 J / (kg·K). The allowable indoor-outdoor temperature difference is set at 5℃.

[0090] The calculation method for the reference frequency is as follows:

[0091] f base =V / V max ×50Hz

[0092] Among them, f base V is the reference frequency for all wall-mounted fans. max This represents the maximum ventilation volume of the fan at 50Hz.

[0093] During the accelerated dehumidification phase, the acceleration ratio is calculated as follows:

[0094] a = 0.6 × (1 - T / T) max )+0.4×(1-D / D max )

[0095] Where 'a' is the acceleration ratio, and 'T' is the current countdown. max The warning threshold is 15 minutes, and D represents the magnitude of the current target absolute humidity difference. max This serves as a reference value for the magnitude of the difference.

[0096] In the acceleration ratio, the countdown timer has a weight of 0.6, and the difference magnitude has a weight of 0.4. A shorter countdown indicates a more urgent dehumidification window, and a smaller difference magnitude indicates a narrower remaining dehumidification space. These two factors are combined to determine the acceleration intensity. Based on the acceleration ratio, the exhaust-side inverter frequency is increased from the baseline inverter frequency, with a maximum increase of 50Hz. Using this technical solution, the frequency increase for accelerated dehumidification has a clear physical basis rather than being arbitrary, and the combined use of the countdown timer and difference magnitude ensures that the acceleration intensity matches the urgency of dehumidification.

[0097] Reference Figure 3 As shown, the IoT-based indoor environmental management system for power distribution stations includes:

[0098] The multi-source sensing module is used to collect data on indoor temperature and humidity of each zone of the substation, outdoor temperature and humidity of the corresponding wall outside each zone, and equipment load current. It calculates the absolute humidity difference between the inside and outside of each zone and takes the minimum value of the absolute humidity difference between the inside and outside of each zone as the target absolute humidity difference between the indoor and outdoor areas of the substation.

[0099] The trend prediction module is used to input the collected data and absolute humidity difference into the time series prediction network and output the temperature and humidity prediction sequence of each zone for future time periods.

[0100] The airflow management decision module is used to determine the heat load distribution based on the load current of each zone's equipment and to allocate the air intake and exhaust roles of the wall fans. It identifies the reversal time based on the target absolute humidity difference sign and the temperature and humidity prediction sequence, and generates management instructions for the wall fans based on the target difference sign and the reversal time.

[0101] The wall-mounted fan control module is used to control the forward and reverse rotation and frequency conversion of multiple wall-mounted fans on the wall of the substation according to the control instructions, so that the sum of the frequency conversion on the air intake side is higher than the sum of the frequency conversion on the exhaust side to maintain a slight positive pressure in the substation.

[0102] In practical implementation, taking a 10kV substation as an example, four wall-mounted fans are arranged diagonally on the substation walls: two on the east wall and two on the west wall, forming a diagonal airflow path. After the multi-source sensing module collects data, it outputs it to the trend prediction module. The trend prediction module outputs the prediction sequence to the airflow governance decision module. The airflow governance decision module generates governance instructions and issues them to the wall-mounted fan governance module for execution, forming a complete governance closed loop of "sensing-prediction-decision-execution".

[0103] The multi-source sensing module deploys temperature and humidity sensors in pairs on both the inner and outer sides of the wall in each partition, so that the sign of the absolute humidity difference reflects the direction of humid air penetration through the wall on that side; the equipment load current is collected from the feeder circuit of each partition; the collected data is filtered by sliding window mean and then output to the trend prediction module.

[0104] In the specific implementation process, one temperature and humidity sensor is deployed on each of the inner and outer sides of the wall in the transformer area on the east side, and one temperature and humidity sensor is deployed on each of the inner and outer sides of the wall in the switch cabinet area on the west side, for a total of four temperature and humidity sensors. This paired deployment ensures that the sign of the absolute humidity difference directly reflects the direction of humid air infiltration through that side of the wall; a positive difference indicates humid air infiltration from indoors to outdoors, and a negative difference indicates humid air infiltration from outdoors to indoors. The collected data is output after being filtered by a sliding window mean filter. The filter window uses 12 sampling points from the past hour, effectively suppressing sensor noise interference with subsequent judgments.

[0105] The trend prediction module executes the time-series prediction network with a preset inference cycle, and outputs an hourly temperature and humidity prediction sequence covering a preset future time period each time. The prediction sequence includes hourly predicted values ​​of the absolute humidity difference, so that the airflow governance decision module can detect the change in the sign of the target difference step by step.

[0106] In practical implementation, the trend prediction module operates on a 5-minute inference cycle, outputting an hourly temperature and humidity prediction sequence covering the next 2 hours, with a prediction time step of 5 minutes and a total of 24 prediction time steps. The prediction sequence includes hourly predicted values ​​of the absolute humidity difference. The airflow management decision module detects changes in the sign of the target difference step by step within each inference cycle. Using this technical solution, the trend prediction module operates on a fixed cycle, ensuring the timeliness of the prediction sequence, and the hourly output of the difference prediction values ​​allows the decision module to identify reversals before they occur.

[0107] The airflow management decision module establishes a four-state control state machine based on the identification results of the target difference sign and the reversal time. When the target difference is positive and no reversal time is identified, the current state is determined to be ventilation and dehumidification state; when the target difference is positive and a reversal time from positive to negative is identified, the current state is determined to be accelerated dehumidification state; when the target difference is negative or zero and no reversal time is identified, the current state is determined to be sealed dehumidification state; when the target difference is negative or zero and a reversal time from negative to positive is identified, the current state is determined to be waiting for ventilation state.

[0108] The airflow management decision module generates corresponding management instructions based on the current state. These instructions include the role identifier, operating direction, and frequency conversion frequency of each wall fan. The airflow management decision module also receives the cumulative running time of each wall fan from the wall fan management module. When at least two zones have a heat load difference lower than a preset heat load balancing threshold, a role reassignment instruction is generated based on the cumulative running time.

[0109] In the specific implementation process, the data structure of the governance command includes three fields: the role identifier (intake / exhaust) of each wall fan, the operating direction (forward / reverse), and the frequency conversion frequency (0 to 50Hz). The airflow governance decision module acquires the current target difference sign and the identification result of the reversal time in each inference cycle and compares it with the state determination result of the previous inference cycle. When the target difference sign or the identification result of the reversal time changes, the four-state control state machine executes a state transition, and the airflow governance decision module generates a new governance command based on the current state after the transition. When only the role reassignment command changes, the four-state control state machine maintains the current state, and the airflow governance decision module updates the role identifier, operating direction, and frequency conversion frequency in the governance command.

[0110] The trigger condition for the role reassignment instruction is: the difference in load current between at least two zones is less than 5A. At this time, the role reassignment instruction is generated based on the cumulative running time of each wall fan, and the wall fan with the shorter cumulative running time is assigned as the air intake role.

[0111] The wall-mounted ventilation fan control module includes:

[0112] The frequency control unit outputs frequency commands to the frequency converter based on the frequency parameters in the management instructions;

[0113] The forward / reverse switching unit switches between forward air intake and reverse air exhaust according to the role identifier in the management instruction and performs electrical interlock protection during the switching process;

[0114] The role adjustment unit records the cumulative runtime of each wall fan and its current role, and feeds it back to the airflow management decision module for role reassignment.

[0115] The system also includes an edge computing gateway, deployed within the substation, which executes the inference of the trend prediction module and the airflow management decision module locally and then directly sends the management instructions to the wall fan management module.

[0116] In practical implementation, the frequency converter control unit outputs frequency commands to the frequency converters of each wall-mounted fan according to the frequency parameters in the governance command. The frequency converters then adjust the motor speed upon receiving the frequency commands. The forward / reverse switching unit switches between forward air intake and reverse air exhaust based on the role identifier in the governance command. During switching, electrical interlock protection is implemented, i.e., the fan stops for 0.5 seconds before reversing to prevent arcing damage to the motor windings during forward / reverse switching. The role adjustment unit records the cumulative runtime and current role of each wall-mounted fan with minute-level precision and feeds back the cumulative runtime to the airflow governance decision module for role reassignment. The edge computing gateway uses an industrial-grade embedded computing platform deployed within the substation. After executing the inference of the trend prediction module and the airflow governance decision module locally, it directly sends the governance commands to the wall-mounted fan governance module, without relying on a cloud server. This technical solution ensures the real-time nature of the governance commands and avoids the impact of network latency on the control effect.

[0117] Example 1:

[0118] Using the summer operating conditions of a 10kV substation as the test scenario, the substation was divided into two areas: the transformer area on the east side and the switchgear area on the west side, with two wall-mounted fans arranged diagonally in each area. The test period was from July to August, selecting typical high-humidity weather with an outdoor temperature of 35°C and a relative humidity of 80%, and continuous monitoring was conducted for 48 hours.

[0119] For ease of description, in this embodiment, the ventilation and dehumidification state, the accelerated dehumidification state, the closed dehumidification state, and the waiting ventilation state are respectively referred to as state A, state B, state C, and state D.

[0120] The operation process of this invention is as follows: Initially, the indoor absolute humidity is 22.5 g / m³, and the outdoor absolute humidity is 18.2 g / m³, with a positive difference (+4.3 g / m³), placing the system in state A. The wall fans operate according to their roles, maintaining a slight positive pressure. At the 6th hour, the time-series prediction network predicts that the reversal point when the difference will turn from positive to negative is 45 minutes away, triggering state B. The system synchronously adjusts the frequency converters on the intake and exhaust sides—gradually decreasing the intake frequency from 45 Hz to 40 Hz and gradually increasing the exhaust frequency from 30 Hz to 40 Hz, reducing the frequency difference from 15 Hz to 0 Hz while maintaining the intake frequency at a level no lower than the exhaust frequency, maximizing dehumidification capacity under slight positive pressure constraints. Accelerated dehumidification stops when the sum of the frequency converters' differences decreases to 0 Hz. Five minutes before the reversal point, all fans are shut down, and the dehumidifier is activated, transitioning to state C. At the 8.5th hour, the system predicts that the reversal point when the difference will turn from negative to positive is 30 minutes away, triggering state D. The system remains off. Once the reversal point is reached, ventilation and slight positive pressure are activated, returning the system to state A.

[0121] The traditional solution (i.e., threshold alarm and fixed-direction ventilation) operates as follows: when indoor humidity exceeds 75%RH, the fan is activated for ventilation, with the fan operating in a fixed direction of east-intake and west-exhaust. Ventilation continues even when outdoor humidity is higher than indoor humidity, causing a large influx of humid outdoor air into the room.

[0122]

[0123] Table 1

[0124] As shown in Table 1, the experimental results show that the present invention uses a four-state control strategy to promptly shut off the fan when the outdoor humidity is higher than the indoor humidity to avoid introducing humid air, and accelerates dehumidification when the dehumidification window is about to close to make full use of the remaining ventilation time. This reduces the maximum indoor humidity by 14 percentage points and reduces the number of condensation events to zero.

[0125] Example 2:

[0126] Within the same substation, the effect of a slight positive pressure on inhibiting the infiltration of humid air through wall gaps was tested. The frequency converter on the air intake side was set to 45Hz, and the frequency converter on the exhaust side was set to 38Hz, maintaining a frequency difference of 7Hz. A miniature humidity probe was installed 5cm away from the wall gap to simultaneously measure the humidity of the main indoor unit and the humidity at the gap.

[0127]

[0128] Table 2

[0129] As shown in Table 2, the experimental results indicate that the humidity at the gap was significantly higher than the indoor humidity when there was no pressure difference, suggesting that outdoor humid air infiltrated into the gap without any organization, forming a localized high humidity condensation point. By maintaining a slight positive pressure of 8 Pa, the humidity at the gap decreased to near the indoor humidity, effectively eliminating the localized high humidity condensation point.

[0130] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention.

Claims

1. A method for monitoring the indoor environment of a power distribution station based on the Internet of Things, characterized in that, Includes the following steps: S1. Collect data on indoor temperature and humidity of each zone of the substation, outdoor temperature and humidity of the corresponding wall outside each zone, and equipment load current. Calculate the absolute humidity difference between the inside and outside of each zone and take the minimum value of the absolute humidity difference between the inside and outside of each zone as the target absolute humidity difference between the indoor and outdoor areas of the substation. S2. Input the collected data and absolute humidity difference into the time series prediction network, and output the temperature and humidity prediction sequence of each zone for future time periods. S3. Determine the heat load distribution and sort the heat loads according to the load current of each zone's equipment. Assign the side wall fans of the zones in the first half of the heat load sorting sequence as exhaust fans and the side wall fans of the zones in the second half of the heat load sorting sequence as intake fans, forming an exhaust side wall fan set and an intake side wall fan set. At the same time, identify the reversal time based on the sign of the target absolute humidity difference and in combination with the temperature and humidity prediction sequence. The reversal time is the moment when the sign of the target difference changes from positive to negative. S4. When the target difference is positive and the reversal moment has not been identified, each wall fan is operated according to the assigned role to make the sum of the frequency converters on the air intake side higher than the sum of the frequency converters on the air exhaust side in order to maintain a slight positive pressure in the substation room. When the moment when the target difference is about to reverse from positive to negative is detected, the dehumidification is accelerated and all wall fans are turned off and the dehumidifier is activated before the reversal moment is reached. S5. When the target difference is negative or zero and the reversal moment is not detected, shut down all wall fans and activate the dehumidifier. When the moment when the target difference is about to turn from negative to positive is detected, the system remains closed until the moment of reversal is reached. Each wall fan is operated according to its assigned role to ensure that the sum of the frequency converters on the air intake side is higher than the sum of the frequency converters on the air exhaust side in order to maintain a slight positive pressure in the substation room.

2. The method for monitoring the indoor environment of a power distribution station based on the Internet of Things according to claim 1, characterized in that, The reversal time is identified based on the sign of the target absolute humidity difference and in conjunction with the temperature and humidity prediction sequence, including: After detecting a change in the sign of the target absolute humidity difference in the temperature and humidity prediction sequence, the target difference sign in subsequent prediction time steps is continuously monitored. Only when the changed target difference sign is continuously maintained for more than the preset confirmation time is it determined to be a valid reversal moment. The countdown is determined based on the time distance between the effective reversal time and the current time. When the countdown is lower than the preset warning threshold, the corresponding control strategy is triggered.

3. The method for monitoring the indoor environment of a power distribution station based on the Internet of Things according to claim 1, characterized in that, This also includes character adjustment A and character adjustment B. The role adjustment A is as follows: the heat load sorting is updated according to a preset inference cycle. When the updated heat load sorting is consistent with the sorting before the change within multiple consecutive preset inference cycles and the cumulative duration exceeds the preset stable duration, the role allocation is switched according to the updated heat load sorting. If the preset stable duration is not exceeded, the original role remains unchanged. The role adjustment B is as follows: when there are at least two zones with heat load differences lower than the preset heat load balance threshold, the roles are reassigned according to the cumulative running time of the corresponding side wall fans, and the wall fans with shorter cumulative running time are assigned as air intake roles. Role adjustment A is used for role adjustment when the heat load sequence changes, and role adjustment B is used for role adjustment when the heat loads of adjacent zones are close. When both role adjustment A and role adjustment B are satisfied, role adjustment B is executed first. The heat load difference is the absolute difference between the heat load characterization values ​​of the two zones, which are determined by the zone equipment load current or the product of the square of the zone equipment load current and the equivalent resistance.

4. The method for monitoring the indoor environment of a power distribution station based on the Internet of Things according to claim 1, characterized in that, Maintaining a slightly positive pressure indoor in the substation includes: The frequency converters on the intake and exhaust sides are adjusted in a closed loop based on the indoor and outdoor pressure difference. The sum of the frequency converters on the intake side is the sum of the frequency converters of all wall fans in the intake side wall fan assembly, and the sum of the frequency converters on the exhaust side is the sum of the frequency converters of all wall fans in the exhaust side wall fan assembly. This maintains the indoor and outdoor pressure difference between 5 and 15 Pa, so that the airflow flows along a directional path from the intake side to the exhaust side and suppresses local condensation caused by unorganized infiltration from wall gaps. When the pressure difference is less than 5 Pa, the sum of the frequency converters on the intake side is increased and the sum of the frequency converters on the exhaust side is decreased. When the pressure difference is greater than 15 Pa, the sum of the frequency converters on the intake side is decreased and the sum of the frequency converters on the exhaust side is increased. It also includes a micro-positive pressure constraint for the accelerated dehumidification stage: when the moment when the target difference is about to reverse from positive to negative is detected, the accelerated dehumidification stage is entered; the sum of the frequency converters on the exhaust side is gradually increased while the sum of the frequency converters on the intake side is simultaneously decreased to accelerate dehumidification, with the constraint that the sum of the frequency converters on the intake side is not lower than the sum of the frequency converters on the exhaust side. When the difference in the sum of the frequency converters shrinks to 0Hz, the accelerated dehumidification is stopped, and all wall fans are turned off and the dehumidifier is activated.

5. The method for monitoring the indoor environment of a power distribution station based on the Internet of Things according to claim 4, characterized in that, Gradually increase the sum of the inverter frequencies on the exhaust side while simultaneously decreasing the sum of the inverter frequencies on the intake side to accelerate dehumidification, including: The heat generation of the equipment is calculated based on the load current of each zone's equipment. The baseline ventilation volume required to maintain a stable indoor temperature in the substation is then calculated based on the current equipment heat generation and the outdoor temperature. The baseline frequency conversion of all wall fans is determined based on the baseline ventilation volume. During the accelerated dehumidification phase, the acceleration ratio is determined by combining the countdown time and the difference between the current target absolute humidity. The acceleration ratio is constrained to have an upper limit of 0Hz, which means that the difference between the current sum of the frequency conversions on the intake side and the current sum of the frequency conversions on the exhaust side is not less than 0Hz. Based on the acceleration ratio, the frequency conversion on the exhaust side is increased and the frequency conversion on the intake side is decreased simultaneously on the basis of the baseline frequency conversion, keeping the sum of the frequency conversions on the intake side not lower than the sum of the frequency conversions on the exhaust side to maintain a slight positive pressure in the substation.

6. An IoT-based indoor environmental management system for power distribution substations, applied to the IoT-based indoor environmental monitoring method for power distribution substations as described in any one of claims 1-5, characterized in that, include: The multi-source sensing module is used to collect data on indoor temperature and humidity of each zone of the substation, outdoor temperature and humidity of the corresponding wall outside each zone, and equipment load current. It calculates the absolute humidity difference between the inside and outside of each zone and takes the minimum value of the absolute humidity difference between the inside and outside of each zone as the target absolute humidity difference between the indoor and outdoor areas of the substation. The trend prediction module is used to input the collected data and absolute humidity difference into the time series prediction network and output the temperature and humidity prediction sequence for each zone in the future period. The airflow management decision module is used to determine the heat load distribution based on the load current of each zone's equipment and to allocate the air intake and exhaust roles of the wall fans. It identifies the reversal time based on the target absolute humidity difference sign and the temperature and humidity prediction sequence, and generates management instructions for the wall fans based on the target difference sign and the reversal time. The wall-mounted fan control module is used to control the forward and reverse rotation and frequency conversion of multiple wall-mounted fans on the wall of the substation according to the control instructions, so that the sum of the frequency conversion on the air intake side is higher than the sum of the frequency conversion on the exhaust side to maintain a slight positive pressure in the substation.

7. The IoT-based indoor environmental management system for power distribution stations according to claim 6, characterized in that, The multi-source sensing module deploys temperature and humidity sensors in pairs on both the inner and outer sides of the wall of each partition, so that the sign of the absolute humidity difference between the inside and outside of each partition reflects the direction of humid air penetration through the corresponding side wall; the equipment load current is collected from the feeder circuit of each partition; the collected data is filtered by the sliding window mean and then output to the trend prediction module.

8. The IoT-based indoor environmental management system for power distribution stations according to claim 6, characterized in that, The trend prediction module executes the time-series prediction network with a preset inference cycle, and outputs an hourly temperature and humidity prediction sequence covering a preset future time period each time. The prediction sequence includes hourly predicted values ​​of the absolute humidity difference, so that the airflow governance decision module can detect the change in the sign of the target difference step by step.

9. The IoT-based indoor environmental management system for power distribution stations according to claim 6, characterized in that, The airflow management decision module establishes a four-state control state machine based on the identification results of the target difference sign and the reversal time. The four-state control state machine includes: a ventilation and dehumidification state when the target difference is positive and the reversal time has not been identified; an accelerated dehumidification state when the target difference is positive and the reversal time from positive to negative has been identified; a closed dehumidification state when the target difference is negative or zero and the reversal time has not been identified; and a waiting ventilation state when the target difference is negative or zero and the reversal time from negative to positive has been identified. The airflow management decision module generates corresponding management instructions based on the current status. The management instructions include the role identification, operating direction and frequency conversion frequency of each wall fan. The airflow management decision module also receives the cumulative running time of each wall fan from the wall fan management module. When there are at least two zones with heat load differences lower than the preset heat load balancing threshold, a role reassignment instruction is generated based on the cumulative running time. The four-state control state machine performs a state transition when the identification result of the target difference sign or reversal moment changes, and updates the role identifier, running direction and frequency conversion frequency in the governance instruction when the role reassignment instruction changes.

10. The IoT-based indoor environmental management system for power distribution stations according to claim 9, characterized in that, The wall-mounted ventilation fan control module includes: The frequency control unit outputs frequency commands to the frequency converter based on the frequency parameters in the management instructions; The forward / reverse switching unit switches between forward air intake and reverse air exhaust according to the role identifier in the management instruction and performs electrical interlock protection during the switching process; The role adjustment unit records the cumulative runtime of each wall fan and its current role, and feeds it back to the airflow management decision module for role reassignment.

Images