Switch cabinet room air conditioning energy-saving control method considering transformer load rate

Abstract

This invention discloses an energy-saving control method for air conditioning in substation switchgear rooms that takes into account transformer load rate. The method collects transformer operation data, indoor and outdoor environmental data, hot spots within the cabinet, condensation detection surfaces, and air conditioning operation parameters to determine the transformer load rate, load rate change rate, load rate change sequence, and current hot spot temperature rise within the cabinet. A coupled prediction model of load rate trend and hot spot temperature rise is established, outputting predicted hot spot temperature rise values. Interval control parameters are determined according to load status, and natural ventilation gating results are generated based on dew point temperature and condensation margin. Cooling, dehumidification, ventilation, or interlocking protection commands are generated based on safety constraints, gating results, and predicted values. Closed-loop correction is performed based on actual hot spot temperature rise, balancing pre-cooling for sudden load increases, natural ventilation to prevent condensation, and safety priority under abnormal operating conditions. This achieves adaptive optimization of equipment safety, energy-saving operation, and control strategies, and supports background alarm transmission, reducing ineffective cooling and false ventilation activation.

Description

Technical Field

[0001] This invention relates to the fields of intelligent operation and maintenance of substations, environmental control of switchgear rooms, energy-saving control of HVAC systems, and thermal risk prediction of power equipment. Specifically, it relates to an energy-saving control method for air conditioning in substation switchgear rooms that takes into account transformer load rate. In particular, it relates to a method that collects and determines transformer operating parameters, indoor environmental parameters of switchgear rooms, hot spot node parameters inside the cabinet, outdoor environmental parameters, condensation judgment surface parameters inside the cabinet, and air conditioning operating parameters. It uses the transformer load rate and its changing trend as the feedforward quantity of the main equipment's thermal load, uses the current temperature rise of hot spots inside the cabinet as the prediction object, and combines load range control, natural ventilation gating, safety constraint judgment, and closed-loop correction to achieve energy-saving control of air conditioning in switchgear rooms. Background Technology

[0002] The switchgear room in a substation houses key power equipment such as circuit breakers, disconnectors, instrument transformers, busbars, cable joints, and secondary control equipment. Typical heat-generating locations within the switchgear include busbar connections, circuit breaker contacts, cable joints, disconnector contacts, and heating terminals. Their temperature rise directly impacts the reliability of conductive connections, the aging rate of insulation components, contact performance, and the long-term operational safety of the switchgear. For unattended substations, environmental control in the switchgear room not only requires maintaining the indoor temperature and humidity to meet equipment operating requirements but also preventing excessively high localized temperatures at hot spots within the cabinets and avoiding condensation formation on the metal surfaces of the switchgear, near busbar supports, near cable joints, or near circuit breaker contacts after low-temperature, high-humidity air enters the switchgear room.

[0003] Existing environmental control schemes for switchgear rooms or power distribution rooms typically control the start and stop of air conditioners, fans, and dehumidifiers based on indoor and outdoor temperatures, humidity, or the environmental conditions of a single area. For example, air conditioning is activated when the indoor temperature exceeds a set value, dehumidification is activated when the indoor humidity exceeds a set value, and ventilation is activated when the outdoor temperature is lower than the indoor temperature. While such schemes can achieve basic environmental regulation, their control is mainly based on indoor ambient temperature and indoor relative humidity. They do not fully consider the feedforward effect of transformer load rate changes on the heat load of the switchgear room, nor do they consider the current temperature rise of hot spots inside the cabinet, such as busbar connections, circuit breaker contacts, and cable joints, as a direct risk to air conditioning control.

[0004] Furthermore, existing technologies also employ predictive models or neural network models for temperature prediction and air conditioning energy-saving control. These models can predict future temperature trends based on historical temperatures or environmental conditions and adjust the air conditioning operation in advance. However, such predictive control is typically geared towards the ambient temperature of ordinary spaces or the operating status of air conditioning systems. The predictive inputs are mostly ambient temperature and humidity, air conditioning operating parameters, or historical temperature data. They do not establish a coupling relationship between transformer load rate, load rate change rate, load rate change sequence, and the current temperature rise of hot spots in the cabinet. Since an increase in transformer load rate will lead to an increase in the heat load of the main equipment, and a sudden increase in load rate often precedes a significant increase in indoor ambient temperature, relying solely on indoor ambient temperature for a passive response can easily result in pre-cooling lag in scenarios of sudden load increases.

[0005] Furthermore, existing natural ventilation energy-saving controls typically determine whether to turn on fans or ventilation windows based on the indoor-outdoor temperature difference and outdoor humidity. When the outdoor ambient temperature is low but the outdoor air dew point is high, directly introducing outdoor air may cause condensation on the low-temperature surfaces inside the switchgear. Especially when the surface temperature of the switchgear cabinet metal surface, the surface near the busbar support, the surface near the cable joint, or the surface near the circuit breaker contacts is low, even if the outdoor temperature meets the cooling conditions, there may still be a risk of condensation due to the high outdoor dew point temperature. Therefore, natural ventilation control based solely on the indoor-outdoor temperature difference and humidity threshold is not suitable for direct application to the complex operating environment of the switchgear room.

[0006] Therefore, it is evident that the existing air conditioning control technology for switchgear rooms has the following shortcomings: First, the control strategy is mainly based on ambient temperature and humidity, and lacks feedforward perception of the main equipment thermal load of transformer load rate, load rate change rate and load rate change sequence. Second, the predicted objects are mostly indoor ambient temperatures, which cannot directly reflect the actual thermal risks of hot spots inside the cabinet, such as busbar connections, circuit breaker contacts, and cable joints. Third, there is a lack of differentiated linkage between the pre-cooling trigger threshold, air conditioning set temperature and humidity threshold, and basic control priority under light load zone, medium load zone, heavy load zone, and load surge zone. Fourth, the natural ventilation control lacks gate control judgment based on indoor dew point temperature, outdoor dew point temperature, and condensation inside the cabinet to determine surface temperature and condensation margin, which leads to a conflict between energy-saving control and equipment safety boundaries. Therefore, there is an urgent need for an energy-saving control method for switchgear room air conditioning, which is based on the load heat source of substation main equipment, the temperature rise risk of hot nodes in the cabinet, and the safety boundary of condensation. Summary of the Invention

[0007] The purpose of this invention is to overcome the problems in the existing technology of air conditioning control in switchgear room mainly relying on indoor and outdoor temperature and humidity, unable to feedforward to sense changes in transformer load, unable to predict and control the temperature rise of hot nodes in the cabinet, natural ventilation has the risk of condensation, and there is a lack of clear control logic between energy-saving control and safety protection. The invention provides an energy-saving control method for air conditioning in substation switchgear room that takes into account transformer load rate.

[0008] The core of this invention lies in first collecting and determining basic control parameters, including transformer operating parameters, switchgear room indoor environmental parameters, hot spot node parameters inside the cabinet, outdoor environmental parameters, condensation judgment surface parameters inside the cabinet, and air conditioning operating parameters; then, determining the transformer load rate, load rate change rate, and load rate change sequence based on the transformer operating parameters, and determining the current hot spot temperature rise inside the cabinet based on the hot spot node parameters inside the cabinet and the indoor ambient temperature; subsequently, establishing a load rate trend-hot spot temperature rise coupled prediction model, outputting the predicted value of the hot spot temperature rise inside the cabinet within a preset future time period; dividing the current control state based on the transformer load rate and load rate change rate, and determining the interval control parameters based on the current control state; further generating natural ventilation gating results through indoor dew point temperature, outdoor dew point temperature, condensation judgment surface temperature inside the cabinet, and condensation margin; finally, combining the safety constraint judgment results, natural ventilation gating results, predicted value of hot spot temperature rise inside the cabinet, current control state, and interval control parameters to generate control commands for the air conditioning, fan, and dehumidification modules, and performing closed-loop correction based on the actual hot spot temperature rise inside the cabinet.

[0009] To achieve the above objectives, the present invention adopts the following technical solution: An energy-saving control method for air conditioning in a substation switchgear room, taking into account transformer load rate, includes the following steps: S1. Collect and determine the basic control parameters, including transformer operating parameters, switchgear room indoor environmental parameters, switchgear hot spot node parameters, outdoor environmental parameters, cabinet condensation judgment surface parameters, and air conditioning operating parameters.

[0010] The main operating parameters of a transformer include its rated capacity, line voltage, line current, top oil temperature, and winding temperature. The load rate of the transformer is determined based on its actual apparent power and rated capacity. The load rate change rate and the load rate change sequence within a preset time window are then determined based on the transformer load rate.

[0011] The indoor environmental parameters of the switchgear room include indoor ambient temperature and indoor relative humidity; the hot spot node parameters inside the cabinet include the temperature of at least one hot spot node inside the cabinet, and the current hot spot temperature rise inside the cabinet is determined based on the hot spot node temperature inside the cabinet and the indoor ambient temperature; the outdoor environmental parameters include outdoor ambient temperature and outdoor relative humidity.

[0012] The parameters for judging condensation inside the cabinet include the surface temperature for judging condensation inside the cabinet, which is the lowest surface temperature inside the cabinet that has been collected, or the surface temperature of the cabinet that has been preset according to the cabinet structure; the air conditioning operating parameters include the air conditioning operating status.

[0013] Specifically, the transformer load rate satisfies β=100×S / Sn, where S is the actual apparent power of the transformer, Sn is the rated capacity of the transformer, and the unit of β is %; the load rate change rate is used to characterize how fast the transformer load rate changes within a preset time window, and can be determined based on the difference between the transformer load rate at the current sampling moment and the transformer load rate at the previous statistical moment or the beginning moment of the sliding time window; the load rate change sequence is composed of the transformer load rates in multiple consecutive sampling periods in chronological order.

[0014] The current hot spot temperature rise inside the cabinet is the difference between the temperature of the hot spot node inside the cabinet and the indoor ambient temperature. When there are multiple hot spot nodes inside the cabinet, the maximum value of the temperature rise of each hot spot node can be used as the current hot spot temperature rise inside the cabinet. Alternatively, the risk equivalent hot spot temperature rise can be calculated based on the equipment risk weight of different hot spot nodes inside the cabinet, and the risk equivalent hot spot temperature rise can be used as the current hot spot temperature rise inside the cabinet.

[0015] S2. Establish a load rate trend-hot spot temperature rise coupled prediction model. Take the transformer load rate, load rate change rate, load rate change sequence, indoor ambient temperature, indoor relative humidity, top oil temperature, current hot spot temperature rise in the cabinet and air conditioning operation status determined in step S1 as the model input, and output the predicted value of hot spot temperature rise in the cabinet within a preset time period.

[0016] Preferably, the load rate trend-hot spot temperature rise coupled prediction model adopts a long short-term memory network model; the model input is no longer limited to ordinary ambient temperature or historical temperature, but introduces transformer load rate, load rate change rate and load rate change sequence as feedforward features of main equipment heat load, introduces the current hot spot temperature rise in the cabinet as the local thermal risk feature of the switchgear, introduces the top oil temperature as the thermal inertia feature of the transformer, and combines the air conditioning operation status to jointly predict the hot spot temperature rise in the cabinet within the next 10 to 30 minutes.

[0017] With the above settings, the model output is not the ordinary indoor temperature, but the future temperature rise trend of the hot spot nodes in the cabinet relative to the indoor ambient temperature; the model input is not simply the ambient temperature and humidity, but simultaneously takes into account the changes in the main equipment load heat source and the heat accumulation state of the hot spot in the cabinet. Therefore, it can identify the risk of hot spot temperature rise caused by a sudden increase in load in advance before the indoor ambient temperature has risen significantly.

[0018] S3. Based on the transformer load rate β and load rate change rate determined in step S1, the current control state is divided into light load zone, medium load zone, heavy load zone and load surge zone.

[0019] The light load zone is defined as a state where the transformer load rate β is less than 30% and the load rate change rate is less than the preset surge threshold; the medium load zone is defined as a state where the transformer load rate β is greater than or equal to 30% and less than or equal to 70%; the heavy load zone is defined as a state where the transformer load rate β is greater than 70%; and the load surge zone is defined as a state where the current transformer load rate β has not reached the heavy load zone but the load rate change rate has reached the preset surge threshold, or a state where the transformer load rate β is predicted to enter the heavy load zone within a preset time period based on extrapolation of the load rate change sequence.

[0020] The significance of introducing the load surge zone in this invention is that when the transformer load rate β is still in the medium load zone but shows a rapid upward trend, the hot spot nodes in the cabinet may be about to enter the stage of rapid temperature accumulation. At this time, if ordinary energy-saving control is still implemented in the medium load zone, it may lead to pre-cooling lag. Therefore, this invention identifies the load surge state by extrapolating the load rate change rate and the load rate change sequence, so that the control strategy changes from passive response to feedforward prediction.

[0021] S4. Determine the interval control parameters based on the current control status obtained in step S3. The interval control parameters include the pre-cooling trigger threshold, the air conditioning set temperature, the humidity threshold, and the basic control priorities for dehumidification, natural ventilation, and cooling.

[0022] Among them, the light load zone corresponds to the first pre-cooling trigger threshold, the medium load zone corresponds to the second pre-cooling trigger threshold, and the heavy load zone and the load surge zone correspond to the third pre-cooling trigger threshold. The first pre-cooling trigger threshold is greater than the second pre-cooling trigger threshold, and the second pre-cooling trigger threshold is greater than the third pre-cooling trigger threshold.

[0023] Preferably, the first pre-cooling trigger threshold is 6℃, the second pre-cooling trigger threshold is 5℃, and the third pre-cooling trigger threshold is 4℃. In the light load area, the basic control priority is dehumidification, natural ventilation, and intermittent cooling, which decreases in that order. In the medium load area, the air conditioning set temperature is corrected according to the formula Tset=32-0.05β, where β is the transformer load rate expressed as a percentage. In the heavy load area and the load surge area, the air conditioning set temperature is lowered to 26℃ to 28℃, and pre-cooling control is started in advance when the predicted temperature rise of the hot spot in the cabinet reaches the third pre-cooling trigger threshold.

[0024] With the above settings, the heavier the load or the faster the load increases, the lower the corresponding pre-cooling trigger threshold and the earlier the pre-cooling action is performed, thus reflecting the linkage between load status, hot spot temperature rise risk and air conditioning control target.

[0025] S5. Calculate the indoor dew point temperature based on the indoor ambient temperature and indoor relative humidity determined in step S1. Calculate the outdoor dew point temperature based on the outdoor ambient temperature and outdoor relative humidity determined in step S1. Determine the dew point temperature of the air to be introduced into the switch cabinet room during natural ventilation as the outdoor dew point temperature. Determine the condensation margin based on the difference between the surface temperature and the outdoor dew point temperature when judging the condensation inside the cabinet.

[0026] The natural ventilation gating results are generated based on the outdoor ambient temperature, indoor ambient temperature, outdoor dew point temperature, indoor dew point temperature, and condensation margin.

[0027] When the outdoor ambient temperature meets the ventilation and cooling conditions, the outdoor dew point temperature is lower than the indoor dew point temperature, and the condensation margin is greater than the preset safety margin, the natural ventilation gating result is that natural ventilation is allowed.

[0028] When the outdoor dew point temperature is not lower than the indoor dew point temperature, or the condensation margin is less than or equal to the preset safety margin, the natural ventilation gating result is to prohibit natural ventilation.

[0029] Specifically, the ventilation and cooling conditions can be such that the outdoor ambient temperature is lower than the indoor ambient temperature by a preset temperature difference, for example, the outdoor ambient temperature is more than 5°C lower than the indoor ambient temperature; the surface temperature for judging condensation inside the cabinet can be the surface temperature collected by temperature sensors installed on the metal surface of the switch cabinet, the surface near the busbar support, the surface near the cable joint, or the surface near the circuit breaker contact; when multiple surface temperature collection points are set, the lowest value among the multiple surface temperatures is used as the surface temperature for judging condensation inside the cabinet.

[0030] With the above settings, natural ventilation is not determined solely by whether the outdoor ambient temperature is low, but is simultaneously constrained by the outdoor dew point temperature, the indoor dew point temperature, and the condensation margin. This reduces the risk of condensation forming on the inner surface of the switch cabinet after low-temperature, high-humidity air enters the switch cabinet room.

[0031] S6. Based on the hot spot node temperature inside the cabinet, indoor ambient temperature, indoor relative humidity, transformer load rate β, winding temperature and air conditioner operating status determined in step S1, generate safety constraint judgment results, and generate control instructions for the air conditioner, fan and dehumidification module based on the safety constraint judgment results, the natural ventilation gating results obtained in step S5, the predicted temperature rise of the hot spot inside the cabinet obtained in step S2, the current control status obtained in step S3 and the interval control parameters obtained in step S4.

[0032] When the safety constraint judgment result is abnormal, a safety interlock protection command is generated, which prohibits energy-saving power reduction control and natural ventilation energy-saving control, and prioritizes the execution of forced cooling, forced dehumidification, standby air conditioning switching or alarm transmission.

[0033] When the safety constraint judgment result is normal and the natural ventilation gating result is to prohibit natural ventilation, a command to prohibit natural ventilation is generated, and dehumidification control or cooling and dehumidification control is executed according to the indoor relative humidity and humidity threshold.

[0034] When the safety constraint judgment result is normal, the natural ventilation gate control result is not to prohibit natural ventilation, and the predicted value of the hot spot temperature rise in the cabinet reaches the pre-cooling trigger threshold corresponding to the current control state, a predictive pre-cooling command is generated, and the air conditioning cooling power is increased or the air conditioning set temperature is reduced.

[0035] When no safety interlock protection command, natural ventilation prohibition command, or pre-cooling prediction command is generated, a dehumidification control command, a natural ventilation energy-saving control command, or a conventional energy-saving cooling command is generated based on the current control status, zone control parameters, indoor relative humidity, and natural ventilation gating results.

[0036] Specifically, the safety constraint judgment result is determined based on the temperature of the hot spot node inside the cabinet, the indoor ambient temperature, the indoor relative humidity, the transformer load rate β, the winding temperature, and the air conditioner operating status. When the temperature of the hot spot node inside the cabinet reaches the first hard constraint threshold, the indoor ambient temperature reaches the second hard constraint threshold, the indoor relative humidity reaches the third hard constraint threshold, the transformer is overloaded or overheated, or the air conditioner malfunctions, the safety constraint judgment result is abnormal and the safety interlock protection is activated.

[0037] When no safety interlock protection command, natural ventilation prohibition command, or pre-cooling prediction command is generated, fuzzy PID control can be performed based on the indoor temperature deviation and the rate of change of temperature deviation to dynamically adjust the PID parameters and output the air conditioner compressor frequency command and fan speed command; wherein, the indoor temperature deviation is the difference between the indoor ambient temperature and the air conditioner set temperature determined in step S4. S7. Execute the control command generated in step S6, and re-collect the temperature of the hot spot node in the cabinet and the indoor ambient temperature after the preset control cycle to determine the actual temperature rise of the hot spot in the cabinet.

[0038] The actual temperature rise of the hot spot inside the cabinet is compared with the predicted temperature rise of the hot spot inside the cabinet output in step S2. When the prediction deviation exceeds the preset deviation threshold, the pre-cooling trigger threshold, preset safety margin, or load rate trend-hot spot temperature rise coupled prediction model parameters corresponding to the current control state are corrected to form a closed-loop control.

[0039] For example, if the actual hot spot temperature rise in the cabinet after a preset time period is higher than the predicted value and the deviation exceeds the preset deviation threshold, it indicates that the load rate trend-hot spot temperature rise coupled prediction model underestimates the risk of hot spot temperature rise. In this case, the pre-cooling trigger threshold corresponding to the current control state can be reduced, the pre-cooling safety factor can be increased, or the input weight of the load rate change rate can be increased. If the actual hot spot temperature rise in the cabinet is significantly lower than the predicted value and the deviation continues to exceed the preset deviation threshold, it indicates that the prediction model may be overly conservative. Under the premise of meeting safety constraints, the energy-saving control tolerance can be appropriately increased to reduce energy waste caused by premature pre-cooling.

[0040] Compared with the prior art, the present invention has at least the following beneficial effects: 1. This invention first collects and determines transformer operating parameters, switchgear room indoor environmental parameters, cabinet hot spot node parameters, outdoor environmental parameters, cabinet condensation judgment surface parameters, and air conditioning operating parameters. Then, based on the above parameters, it performs prediction, zoning, gating, arbitration, and correction, so that each control step has a clear data source and logical connection relationship, avoiding the passivity of simply controlling the start and stop of air conditioning based on indoor temperature and humidity.

[0041] 2. This invention uses the transformer load rate β, load rate change rate, and load rate change sequence as feedforward quantities of the main equipment's thermal load and introduces them into the air conditioning control process of the switchgear room. This can identify the thermal risks caused by a sudden increase in load before the indoor ambient temperature rises significantly. Compared with schemes that simply rely on indoor temperature or indoor and outdoor temperature and humidity control, this invention has stronger feedforward and targeted capabilities.

[0042] 3. This invention uses the temperature rise of hot spots in the cabinet, such as busbar connections, circuit breaker contacts, cable joints, disconnector contacts, and heating connection terminals, as the prediction target, rather than just predicting the ordinary indoor ambient temperature. This allows the air conditioning control target to directly correspond to the safety risks of the switchgear equipment, and can more accurately suppress local hot spot overheating.

[0043] 4. This invention sets different pre-cooling trigger thresholds, air conditioning set temperature and humidity thresholds, and basic control priorities for light load zone, medium load zone, heavy load zone, and load surge zone, respectively. It also makes the first pre-cooling trigger threshold greater than the second pre-cooling trigger threshold, and the second pre-cooling trigger threshold greater than the third pre-cooling trigger threshold, so as to realize zoned control where the pre-cooling is earlier as the load is heavier. This avoids the ineffective cooling of traditional constant temperature control during light load periods and also avoids the pre-cooling lag when the load surges.

[0044] 5. This invention adds the determination of condensation in the cabinet by surface temperature, indoor dew point temperature, outdoor dew point temperature and condensation margin, and generates natural ventilation gate control results based on these. Natural ventilation is only allowed when the outdoor ambient temperature meets the ventilation and cooling conditions, the outdoor dew point temperature is lower than the indoor dew point temperature and the condensation margin is greater than the preset safety margin. This can prevent low temperature and high humidity air from entering the switch cabinet room and causing condensation, thus improving the safety of natural ventilation energy-saving control.

[0045] 6. This invention generates control commands for the air conditioner, fan, and dehumidification module based on the safety constraint judgment results, natural ventilation gate control results, predicted temperature rise of hot spots inside the cabinet, current control status, and interval control parameters. This ensures that there are clear triggering conditions and execution sequences between safety interlock protection, prohibition of natural ventilation, predictive pre-cooling, dehumidification control, natural ventilation energy-saving control, and conventional energy-saving refrigeration, so that energy-saving control always conforms to the equipment safety boundary.

[0046] 7. This invention uses feedback on the deviation between the actual temperature rise of hot spots inside the cabinet and the predicted value to perform closed-loop correction on the parameters of the pre-cooling trigger threshold, preset safety margin, or load rate trend-hot spot temperature rise coupled prediction model, so that the control method can adapt to the thermal response differences under different substations, different switchgear layouts, different air conditioning capacities, and different seasonal environmental conditions. Attached Figure Description

[0047] Figure 1 Flowchart of energy-saving control method for air conditioning in substation switchgear room taking into account transformer load rate; Figure 2 This is a hardware architecture diagram of the air conditioning energy-saving control system for the switchgear room in this invention; Figure 3 This is a schematic diagram of the load rate trend-hot spot temperature rise coupled prediction model structure in this invention; Figure 4 This is a schematic diagram of the natural ventilation gating logic in this invention; Figure 5 This is a schematic diagram of the control instruction generation logic in this invention. Detailed Implementation

[0048] The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings and embodiments. It should be understood that the following embodiments are only used to explain the technical solution of the present invention and are not intended to limit the scope of protection of the present invention. In the absence of conflict, the technical features in this embodiment can be combined with each other.

[0049] This embodiment provides an energy-saving control method for air conditioning in substation switchgear rooms that takes into account transformer load rate. It is applicable to switchgear rooms, distribution rooms, indoor substation equipment rooms, or similar power equipment rooms in 110kV, 220kV, and other voltage level substations. The controlled objects in this embodiment include intelligent variable frequency air conditioners, natural ventilation fans, dehumidification modules, standby air conditioners, and the substation background monitoring system. This embodiment uses an edge computing unit to perform basic control parameter acquisition and determination, load rate trend-hotspot temperature rise coupling prediction, current control state division, interval control parameter determination, natural ventilation gating, control command generation, control execution, and closed-loop correction.

[0050] like Figure 1 As shown, the method in this embodiment includes the following steps: S1. Collect and determine the basic control parameters; S2. Establish a load rate trend-hot spot temperature rise coupled prediction model based on the control basic parameters and output the predicted value of hot spot temperature rise in the cabinet. S3. Divide the current control state according to the transformer load rate β and the load rate change rate; S4. Determine the interval control parameters based on the current control status; S5. Generate natural ventilation gating results based on indoor and outdoor dew point temperatures and condensation margins; S6. Generate control commands for the air conditioning, fan and dehumidification modules based on the safety constraint judgment results, natural ventilation gate control results, predicted temperature rise of hot spots in the cabinet, current control status and interval control parameters; S7. Execute control commands and perform closed-loop correction based on the deviation between the actual temperature rise of the hot spot inside the cabinet and the predicted temperature rise of the hot spot inside the cabinet. After the step is completed, the next control cycle begins.

[0051] like Figure 2 As shown, this embodiment provides an energy-saving control system for air conditioning in a switchgear room, including a data acquisition layer, a communication transmission layer, an edge computing control layer, and an execution and backend layer. The data acquisition layer is used to collect transformer operating parameters, indoor environmental parameters of the switchgear room, hot spot node parameters inside the cabinet, outdoor environmental parameters, condensation judgment surface parameters inside the cabinet, and air conditioning operating parameters. The communication transmission layer is used to transmit data from the data acquisition layer to the edge computing control layer. The edge computing control layer is used to execute the control method of this embodiment. The execution and backend layer is used to receive control commands and execute air conditioning adjustment, natural ventilation, dehumidification, standby air conditioning switching, or alarm transmission.

[0052] The data acquisition layer includes current transformers, voltage transformers, energy meters, or substation backend interfaces for acquiring the actual apparent power of the transformer; transformer temperature controllers or integrated online monitoring devices for acquiring top oil temperature and winding temperature; indoor temperature and humidity sensors for acquiring indoor ambient temperature and relative humidity; fiber optic temperature sensors, infrared temperature sensors, or contact temperature sensors for acquiring the temperature of hot spots inside the cabinet; outdoor temperature and humidity sensors for acquiring outdoor ambient temperature and relative humidity; surface temperature sensors for acquiring condensation inside the cabinet and determining surface temperature; and an air conditioning controller interface for acquiring the air conditioning operating status. The communication transmission layer can use RS485 bus, Ethernet, wireless communication modules, or combinations thereof. The edge computing control layer includes edge computing units. The execution and backend layer includes intelligent variable frequency air conditioners, natural ventilation fans, dehumidification modules, standby air conditioners, and substation backend monitoring systems.

[0053] In step S1, the edge computing unit collects and determines the basic control parameters; the basic control parameters include transformer operating parameters, switchgear room indoor environmental parameters, cabinet hot spot node parameters, outdoor environmental parameters, cabinet condensation judgment surface parameters, and air conditioning operating parameters.

[0054] Transformer operating parameters include rated capacity, line voltage, line current, top oil temperature (Toil), and winding temperature (Tw). The edge computing unit calculates the transformer load factor β based on the actual apparent power S and rated capacity Sn, satisfying β = 100% × S / Sn, where β is expressed as a percentage (%). For example, when the actual apparent power S is 0.6 times the rated capacity Sn, the transformer load factor β is 60%. The edge computing unit determines the load rate change rate and the load rate change sequence within a preset time window based on the transformer load rate β. Specifically, the transformer load rate at the current sampling moment is denoted as β(t), the transformer load rate at the start of the sliding time window is denoted as β(t-τ), and the load rate change rate is denoted as γ. Then, γ can be expressed as γ=β(t)-β(t-τ)] / τ, where τ is the length of the sliding time window. The transformer load rates β of multiple consecutive sampling periods within the preset time window are arranged in chronological order to form a load rate change sequence, denoted as βseq. Through the load rate change rate γ and the load rate change sequence βseq, it can be determined whether the transformer load is in a rapidly increasing state.

[0055] The indoor environmental parameters of the switchgear room include indoor ambient temperature Tin and indoor relative humidity Hin. The indoor ambient temperature Tin and indoor relative humidity Hin can be collected by multiple temperature and humidity sensors arranged in the middle of the switchgear room, near the switchgear side and near the air conditioning return air vent, and the average value, weighted average value or representative measurement point value can be used as indoor environmental parameters for control.

[0056] The parameters of the hot spot nodes inside the cabinet include the temperature of at least one hot spot node inside the cabinet; the hot spot nodes inside the cabinet include at least one of the following: busbar connection, circuit breaker contact, cable joint, disconnector contact, and heating connection terminal inside the cabinet; in this embodiment, fiber optic temperature sensors are installed at the busbar connection, near the circuit breaker contact, and near the cable joint in each row of switch cabinets to collect the temperatures of multiple hot spot nodes inside the cabinet, which are denoted as Th1, Th2, ..., Thn respectively; the edge computing unit determines the current hot spot temperature rise inside the cabinet based on the temperature of the hot spot nodes inside the cabinet and the indoor ambient temperature Tin. Specifically, the temperature rise of each hot spot node in the cabinet is ΔTh_i = Th_i - Tin; the current hot spot temperature rise in the cabinet can be the maximum value among the temperature rises of multiple hot spot nodes in the cabinet, i.e., ΔTh_cur = β = max{ΔTh_i}; or the risk equivalent hot spot temperature rise can be calculated according to the equipment risk weight of different hot spot nodes in the cabinet, and the risk equivalent hot spot temperature rise can be used as the current hot spot temperature rise in the cabinet; using the maximum value or the risk equivalent value as the current hot spot temperature rise in the cabinet can make the control logic prioritize responding to hot spot nodes with higher risks.

[0057] Outdoor environmental parameters include outdoor ambient temperature Tout and outdoor relative humidity Hout; outdoor ambient temperature Tout and outdoor relative humidity Hout can be collected by temperature and humidity sensors installed near the air inlet of the switch cabinet room, at the outdoor environmental monitoring point of the substation, or on the shaded side of the building's exterior wall.

[0058] The surface parameters for judging condensation inside the cabinet include the surface temperature Ts for judging condensation inside the cabinet. The surface temperature Ts for judging condensation inside the cabinet can be the temperature collected by a surface temperature sensor installed on the metal surface of the switch cabinet, the surface near the busbar support, the surface near the cable joint, or the surface near the circuit breaker contact. When multiple surface temperature collection points are set, the lowest value among the multiple surface temperatures is taken as the surface temperature Ts for judging condensation inside the cabinet. For switch cabinets without surface temperature sensors, the cabinet surface temperature can also be preset according to the switch cabinet structure, indoor ambient temperature, and cabinet thermal inertia, and this preset cabinet surface temperature can be taken as the surface temperature Ts for judging condensation inside the cabinet.

[0059] Air conditioner operating parameters include the air conditioner's operating status. The air conditioner's operating status includes at least one of the following: air conditioner on, off, cooling, dehumidifying, and fault status. Furthermore, air conditioner operating parameters may also include the air conditioner's set temperature, return air temperature, compressor operating frequency, fan speed, dehumidification status, and power consumption. These operating parameters can be collected through the air conditioner controller's communication interface.

[0060] like Figure 3As shown, in step S2, the edge computing unit establishes a load rate trend-hot spot temperature rise coupled prediction model. The model takes the transformer load rate β, load rate change rate γ, load rate change sequence βseq, indoor ambient temperature Tin, indoor relative humidity Hin, top oil temperature Toil, current hot spot temperature rise ΔTh_cur in the cabinet, and air conditioner operating status as inputs, and outputs the predicted value of hot spot temperature rise in the cabinet within a preset time period, denoted as ΔTh_pre.

[0061] In this embodiment, the load rate trend-hotspot temperature rise coupled prediction model can adopt a long short-term memory network model. During modeling, the edge computing unit can first perform outlier removal, smoothing filtering, and normalization on the input data, and then construct a time series input according to a preset time window. After receiving the time series input, the long short-term memory network model outputs the predicted value ΔTh_pre of the hotspot temperature rise in the cabinet within the next 10 to 30 minutes. Unlike ordinary indoor temperature prediction, the prediction object in this embodiment is the temperature rise of the hotspot node in the cabinet relative to the indoor ambient temperature, rather than the ordinary indoor ambient temperature. The model input simultaneously includes the transformer load rate β, the load rate change rate γ, and the load rate change sequence βseq. Therefore, it can identify the risk of hotspot temperature rise caused by a sudden increase in load before the indoor ambient temperature rises significantly.

[0062] In step S3, the edge computing unit divides the current control state according to the transformer load rate β and load rate change rate γ determined in step S1; the current control state includes light load zone, medium load zone, heavy load zone and load surge zone.

[0063] Specifically, when β < 30% and γ is less than the preset surge threshold, the current control state is determined to be a light load zone; when 30% ≤ β ≤ 70% and the load surge criterion is not triggered, the current control state is determined to be a medium load zone; when β > 70%, the current control state is determined to be a heavy load zone; when β ≤ 70% but γ reaches the preset surge threshold, or when the load rate change sequence βseq is extrapolated to predict that β will be greater than 70% in the future preset period, the current control state is determined to be a load surge zone.

[0064] The load surge zone setting is used to handle the transition risk of the transformer load rapidly entering the heavy load zone from the medium load zone. For example, if the current transformer load rate β is 65%, which has not yet reached the heavy load zone, but the load rate change rate γ reaches the preset surge threshold, or if it is judged by extrapolation based on the load rate change sequence βseq that β will exceed 70% in the next 10 minutes, then the edge computing unit will determine the current control state as the load surge zone in advance, so that the subsequent step S4 can determine a lower pre-cooling trigger threshold, and step S6 can generate a predictive pre-cooling command when the conditions are met.

[0065] In step S4, the edge computing unit determines the interval control parameters based on the current control state obtained in step S3. The interval control parameters include the pre-cooling trigger threshold, the air conditioning set temperature, the humidity threshold, and the basic control priorities for dehumidification, natural ventilation, and cooling.

[0066] In this embodiment, the light load zone corresponds to the first pre-cooling trigger threshold, which is preferably 6°C; the medium load zone corresponds to the second pre-cooling trigger threshold, which is preferably 5°C; the heavy load zone and the load surge zone correspond to the third pre-cooling trigger threshold, which is preferably 4°C; the first pre-cooling trigger threshold is greater than the second pre-cooling trigger threshold, and the second pre-cooling trigger threshold is greater than the third pre-cooling trigger threshold, so that the heavier the load or the faster the load rises, the lower the pre-cooling trigger threshold and the earlier the pre-cooling action.

[0067] In the light-load area, the basic control priority is dehumidification, natural ventilation, and intermittent cooling in that order, decreasing sequentially. This basic control priority is used for routine energy-saving control when the safety interlock protection command, the natural ventilation prohibition command, and the predicted pre-cooling command are not triggered in step S6. Specifically, in the light-load area, if the indoor relative humidity Hin reaches the humidity threshold, a dehumidification control command is generated first. If the indoor relative humidity Hin does not reach the humidity threshold and the natural ventilation gating result is that natural ventilation is allowed, a natural ventilation energy-saving control command is generated first. If the dehumidification and natural ventilation conditions are not met, a routine energy-saving cooling command is generated based on the temperature deviation between the indoor ambient temperature Tin and the air conditioning set temperature Tset.

[0068] In the medium load area, the air conditioning set temperature Tset is dynamically corrected according to the transformer load rate β, preferably Tset=32-0.05β; for example, when β is 40%, Tset is 30℃; when β is 60%, Tset is 29℃; through this set temperature correction rule, the air conditioning set temperature can be appropriately reduced as the transformer load rate increases, thereby improving the ability to suppress hot spot temperature rise.

[0069] In the heavy load area and the load surge area, the air conditioner set temperature is lowered to 26°C to 28°C; when the predicted value of the hot spot temperature rise ΔTh_pre output in step S2 reaches the third pre-cooling trigger threshold, the edge computing unit generates a predicted pre-cooling command in step S6 to increase the air conditioner cooling power or reduce the air conditioner set temperature, so as to form a cooling margin before the temperature of the hot spot node in the cabinet reaches the hard constraint threshold.

[0070] To ensure a clear correspondence between the interval control parameters and the predicted pre-cooling action, in step S6, the predicted temperature rise value ΔTh_pre of the hot spot inside the cabinet in the light load zone is compared with the first pre-cooling trigger threshold; the predicted temperature rise value ΔTh_pre of the hot spot inside the cabinet in the medium load zone is compared with the second pre-cooling trigger threshold; and the predicted temperature rise value ΔTh_pre of the hot spot inside the cabinet in the heavy load zone and the load surge zone is compared with the third pre-cooling trigger threshold.

[0071] like Figure 4 As shown, in step S5, the edge computing unit calculates the indoor dew point temperature Tdin based on the indoor ambient temperature Tin and indoor relative humidity Hin determined in step S1, and calculates the outdoor dew point temperature Tdout based on the outdoor ambient temperature Tout and outdoor relative humidity Hout determined in step S1; the air to be entered into the switch cabinet room during natural ventilation is outdoor air, so the dew point temperature of the air to be entered into the switch cabinet room during natural ventilation is determined as the outdoor dew point temperature Tdout.

[0072] The edge computing unit determines the condensation margin M based on the difference between the condensation judgment surface temperature Ts inside the cabinet and the outdoor dew point temperature Tdout, i.e., M = Ts - Tdout. The condensation margin M is used to characterize the risk margin of condensation forming on the condensation judgment surface inside the cabinet after outdoor air enters the switch cabinet room. The larger M is, the greater the margin that the condensation judgment surface temperature inside the cabinet is higher than the outdoor dew point temperature, and the lower the condensation risk. The smaller M is, the closer the condensation judgment surface temperature inside the cabinet is to or lower than the outdoor dew point temperature, and the higher the condensation risk.

[0073] The edge computing unit generates a natural ventilation gating result based on the outdoor ambient temperature Tout, indoor ambient temperature Tin, outdoor dew point temperature Tdout, indoor dew point temperature Tdin, and condensation margin M. When the outdoor ambient temperature Tout meets the ventilation and cooling conditions, the outdoor dew point temperature Tdout is lower than the indoor dew point temperature Tdin, and the condensation margin M is greater than the preset safety margin M0, the natural ventilation gating result is to allow natural ventilation. The ventilation and cooling conditions can be Tout≤Tin-ΔT0, where ΔT0 is a preset temperature difference, such as 5℃. When the outdoor dew point temperature Tdout is not lower than the indoor dew point temperature Tdin, or the condensation margin M is less than or equal to the preset safety margin M0, the natural ventilation gating result is to prohibit natural ventilation.

[0074] Therefore, natural ventilation is not determined solely by whether the outdoor ambient temperature is lower than the indoor ambient temperature. Instead, it is constrained by the outdoor dew point temperature, the indoor dew point temperature, and the condensation margin. This prevents low-temperature, high-humidity air from entering the switch cabinet room and causing condensation on the metal surfaces of the cabinet, the surfaces near the busbar supports, the surfaces near the cable joints, or the surfaces near the circuit breaker contacts.

[0075] like Figure 5As shown, in step S6, the edge computing unit generates control commands for the air conditioner, fan, and dehumidification module. The generation of the control commands is based on the following: the temperature of the hot spot node in the cabinet, the indoor ambient temperature Tin, the indoor relative humidity Hin, the transformer load rate β, the winding temperature Tw, and the air conditioner operating status determined in step S1; the natural ventilation gating result obtained in step S5; the predicted value of the hot spot temperature rise ΔTh_pre obtained in step S2; the current control status obtained in step S3; and the interval control parameters obtained in step S4.

[0076] The edge computing unit first generates a safety constraint judgment result based on the hot spot node temperature inside the cabinet, the indoor ambient temperature Tin, the indoor relative humidity Hin, the transformer load rate β, the winding temperature Tw, and the air conditioning operating status. When the hot spot node temperature inside the cabinet reaches the first hard constraint threshold, the indoor ambient temperature Tin reaches the second hard constraint threshold, the indoor relative humidity Hin reaches the third hard constraint threshold, the transformer experiences overload or over-temperature, or the air conditioning equipment malfunctions, the safety constraint judgment result is abnormal. At this time, the edge computing unit generates a safety interlock protection command, prohibiting energy-saving power reduction control and natural ventilation energy-saving control, and prioritizing the execution of forced cooling, forced dehumidification, backup air conditioning switching, or alarm transmission. If the main air conditioning fails and forced cooling cannot be executed, the backup air conditioning is switched first or an alarm is reported to the substation's backend monitoring system.

[0077] When the safety constraint judgment result is normal and the natural ventilation gating result is to prohibit natural ventilation, the edge computing unit generates a command to prohibit natural ventilation, controls the natural ventilation fan to remain closed, and performs dehumidification control or cooling and dehumidification control according to the indoor relative humidity Hin and the humidity threshold determined in step S4. If the predicted value of the hot spot temperature rise ΔTh_pre in the cabinet reaches the pre-cooling trigger threshold corresponding to the current control state at the same time as the natural ventilation gating result prohibits natural ventilation, the cooling and dehumidification control can be performed by lowering the air conditioner set temperature or increasing the air conditioner cooling power to simultaneously meet the requirements of condensation risk control and hot spot temperature rise suppression.

[0078] When the safety constraint judgment result is normal, the natural ventilation gating result is that natural ventilation is allowed, and the predicted value of the hot spot temperature rise ΔTh_pre in the cabinet reaches the pre-cooling trigger threshold corresponding to the current control state, the edge computing unit generates a predicted pre-cooling command and increases the air conditioning cooling power or decreases the air conditioning set temperature. For example, in the heavy load area or the load surge area, when ΔTh_pre reaches the third pre-cooling trigger threshold, the edge computing unit can lower the air conditioning set temperature to 26°C to 28°C and increase the compressor operating frequency and fan speed.

[0079] When no safety interlock protection command, natural ventilation prohibition command, or predictive pre-cooling command is generated, the edge computing unit generates dehumidification control command, natural ventilation energy-saving control command, or conventional energy-saving cooling command based on the current control status, zone control parameters, indoor relative humidity Hin, and natural ventilation gating results. For example, in the light load zone, if the indoor relative humidity Hin reaches the humidity threshold, a dehumidification control command is generated; if the indoor relative humidity Hin does not reach the humidity threshold and the natural ventilation gating result allows natural ventilation, a natural ventilation energy-saving control command is generated; if the natural ventilation conditions are not met and the indoor ambient temperature Tin needs to be adjusted, a conventional energy-saving cooling command is generated; in the medium load zone, the edge computing unit performs conventional energy-saving cooling adjustment based on the air conditioning set temperature Tset; in the heavy load zone and the load surge zone, the edge computing unit prioritizes the suppression of hot spot temperature rise in the cabinet.

[0080] When no safety interlock protection command, natural ventilation prohibition command, or pre-cooling prediction command is generated, the edge computing unit can also perform fuzzy PID control based on the indoor temperature deviation e and the temperature deviation change rate ec. The indoor temperature deviation e is the difference between the indoor ambient temperature Tin and the air conditioning set temperature Tset determined in step S4, i.e., e = Tin - Tset. The temperature deviation change rate ec is de / dt. The fuzzy PID controller uses e and ec as inputs to dynamically adjust the PID parameters Kp, Ki, and Kd, and outputs the air conditioning compressor frequency command and fan speed command. In this way, the frequent start-stop of the compressor can be reduced, and the smoothness of air conditioning control can be improved.

[0081] In step S7, the edge computing unit executes the control command generated in step S6, and after a preset control cycle, re-collects the temperature of the hot spot node inside the cabinet and the indoor ambient temperature to determine the actual temperature rise of the hot spot inside the cabinet. The actual temperature rise of the hot spot inside the cabinet can be determined by the difference between the re-collected temperature of the hot spot node inside the cabinet and the re-collected indoor ambient temperature. When there are multiple hot spot nodes inside the cabinet, the method for determining the actual temperature rise of the hot spot inside the cabinet is consistent with the method for determining the current temperature rise of the hot spot inside the cabinet in step S1.

[0082] The edge computing unit compares the actual hot spot temperature rise in the cabinet with the predicted hot spot temperature rise value ΔTh_pre output in step S2 to obtain the prediction deviation. When the prediction deviation exceeds the preset deviation threshold, the edge computing unit corrects the pre-cooling trigger threshold, preset safety margin, or load rate trend-hot spot temperature rise coupled prediction model parameters corresponding to the current control state.

[0083] Specifically, when the prediction deviation is mainly manifested as the actual hot spot temperature rise in the cabinet exceeding the predicted hot spot temperature rise value ΔTh_pre, it indicates that the load rate trend-hot spot temperature rise coupled prediction model underestimates the risk of hot spot temperature rise. The edge computing unit should reduce the pre-cooling trigger threshold corresponding to the current control state, or adjust the parameters of the load rate trend-hot spot temperature rise coupled prediction model. When the prediction deviation is related to the increase in indoor relative humidity after natural ventilation, insufficient condensation margin, or misjudgment of natural ventilation gate control, the edge computing unit should increase the preset safety margin M0. When the actual hot spot temperature rise in the cabinet is lower than the predicted hot spot temperature rise value ΔTh_pre and the prediction deviation continues to exceed the preset deviation threshold, it indicates that the prediction model may be overly conservative. The edge computing unit can appropriately increase the energy-saving control tolerance under the premise of meeting safety constraints to reduce energy waste caused by premature pre-cooling.

[0084] In this embodiment, the edge computing unit can be an industrial control computer, an embedded controller, a smart gateway, or a substation integrated online monitoring device with data processing capabilities; the intelligent variable frequency air conditioner can receive control commands for the air conditioner's set temperature, compressor frequency, and fan speed through a communication interface; the natural ventilation fan can be installed at the air inlet or outlet of the switch cabinet room and can be used in conjunction with a dust filter module, a backflow preventer, or an electric louver; the dehumidification module can be integrated into the air conditioner or can be a standalone dehumidifier; the backup air conditioner is used to provide redundant cooling capacity in the event of a main air conditioner failure or safety lockout protection.

[0085] This embodiment integrates the transformer load rate β, load rate change rate, load rate change sequence, current hot spot temperature rise in the cabinet, predicted hot spot temperature rise in the cabinet, current control status, interval control parameters, natural ventilation gating results, and safety constraint judgment results into the same control closed loop through the above steps. This makes the air conditioning control of the switchgear room no longer a simple temperature and humidity threshold start-stop control, nor a simple superposition of ordinary temperature prediction and fuzzy PID regulation. Instead, it forms a dedicated air conditioning energy-saving control method for changes in the heat source of the main equipment in the substation, the safety risks of hot spots in the cabinet, and the safety boundary of condensation.

[0086] The working principle of this technical solution is as follows: The edge computing unit uses the transformer operating status as the feedforward source of the heat load change in the switch cabinet room, takes the temperature rise of the hot spot node in the cabinet as the direct control object of the equipment thermal risk, and takes the indoor and outdoor dew point temperature and the condensation judgment surface temperature in the cabinet as the natural ventilation safety boundary. It first continuously acquires the transformer load rate, load rate change rate, load rate change sequence, indoor ambient temperature and humidity, hot spot node temperature in the cabinet, outdoor ambient temperature and humidity, condensation judgment surface temperature in the cabinet, and air conditioning operating status through the data acquisition layer. Based on the above data, predict the temperature rise of hot spots inside the cabinet within the preset time period in the future, and determine differentiated pre-cooling trigger thresholds, air conditioning set temperature and humidity thresholds and basic control priorities according to light load area, medium load area, heavy load area and load surge area; Then, the condensation margin is used to determine whether natural ventilation is allowed. Control arbitration is carried out between the safety constraint judgment result, the natural ventilation gate control result, the predicted value of the hot spot temperature rise in the cabinet and the interval control parameters. Safety interlock protection is given priority, followed by prohibition of natural ventilation, predictive pre-cooling, dehumidification control, natural ventilation energy-saving control or conventional energy-saving cooling. Finally, the control commands are output to the intelligent variable frequency air conditioner, natural ventilation fan, dehumidification module, standby air conditioner and substation background monitoring system, and in the next control cycle, the prediction model parameters, pre-cooling trigger threshold or preset safety margin are corrected by the deviation between the actual hot spot temperature rise in the cabinet and the predicted value. This forms a closed-loop control process of "load feedforward prediction - risk zone control - ventilation and anti-condensation gate control - safety priority arbitration - execution feedback correction", realizing the transformation of switchgear room environmental regulation from a single temperature and humidity threshold start-stop to adaptive energy-saving control oriented towards main equipment load changes, cabinet hot spot risks and condensation safety boundaries.

[0087] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. A method for energy-saving control of air conditioning in switchgear room considering transformer load rate, characterized in that, Includes the following steps: S1. Collect and determine the basic control parameters, including transformer operating parameters, switchgear room indoor environmental parameters, switchgear hot spot node parameters, outdoor environmental parameters, switchgear condensation judgment surface parameters, and air conditioning operating parameters. The main operating parameters of a transformer include its rated capacity, line voltage, line current, top oil temperature, and winding temperature. The load rate of the transformer is determined based on its actual apparent power and rated capacity. The load rate change rate and the load rate change sequence within a preset time window are then determined based on the transformer load rate. The indoor environmental parameters of the switchgear room include indoor ambient temperature and indoor relative humidity; The parameters of the hot spot nodes inside the cabinet include the temperature of at least one hot spot node inside the cabinet, and the current temperature rise of the hot spot inside the cabinet is determined based on the temperature of the hot spot node inside the cabinet and the indoor ambient temperature. Outdoor environmental parameters include outdoor ambient temperature and outdoor relative humidity; The parameters for judging condensation inside the cabinet include the surface temperature for judging condensation inside the cabinet. The surface temperature for judging condensation inside the cabinet is the lowest surface temperature inside the cabinet that has been collected, or the surface temperature of the cabinet that has been preset according to the cabinet structure. Air conditioner operating parameters include the air conditioner's operating status; S2. Establish a load rate trend-hot spot temperature rise coupled prediction model. Take the transformer load rate, load rate change rate, load rate change sequence, indoor ambient temperature, indoor relative humidity, top oil temperature, current hot spot temperature rise in the cabinet and air conditioning operation status determined in step S1 as the model input, and output the predicted value of hot spot temperature rise in the cabinet within the future preset time period. S3. Based on the transformer load rate and load rate change rate determined in step S1, the current control state is divided into light load zone, medium load zone, heavy load zone and load surge zone. The load surge zone is the state where the current transformer load rate has not reached the heavy load zone but the load rate change rate reaches the preset surge threshold, or the transformer load rate β is predicted to enter the heavy load zone in the future preset period based on the load rate change sequence extrapolation. S4. Determine the interval control parameters based on the current control status obtained in step S3. The interval control parameters include the pre-cooling trigger threshold, the air conditioning set temperature, the humidity threshold, and the basic control priorities of dehumidification, natural ventilation, and cooling. Among them, the light load zone corresponds to the first pre-cooling trigger threshold, the medium load zone corresponds to the second pre-cooling trigger threshold, and the heavy load zone and the load surge zone correspond to the third pre-cooling trigger threshold. The first pre-cooling trigger threshold is greater than the second pre-cooling trigger threshold, and the second pre-cooling trigger threshold is greater than the third pre-cooling trigger threshold. S5. Calculate the indoor dew point temperature based on the indoor ambient temperature and indoor relative humidity determined in step S1. Calculate the outdoor dew point temperature based on the outdoor ambient temperature and outdoor relative humidity determined in step S1. Determine the dew point temperature of the air to be entered into the switch cabinet room during natural ventilation as the outdoor dew point temperature. Determine the condensation margin based on the difference between the surface temperature and the outdoor dew point temperature when judging the condensation inside the cabinet. The natural ventilation gating results are generated based on the outdoor ambient temperature, indoor ambient temperature, outdoor dew point temperature, indoor dew point temperature, and condensation margin. When the outdoor ambient temperature meets the ventilation and cooling conditions, the outdoor dew point temperature is lower than the indoor dew point temperature, and the condensation margin is greater than the preset safety margin, the natural ventilation gating result is that natural ventilation is allowed. When the outdoor dew point temperature is not lower than the indoor dew point temperature, or the condensation margin is less than or equal to the preset safety margin, the natural ventilation gating result is to prohibit natural ventilation. S6. Based on the hot spot node temperature inside the cabinet, indoor ambient temperature, indoor relative humidity, transformer load rate, winding temperature and air conditioner operating status determined in step S1, generate safety constraint judgment results, and generate control instructions for the air conditioner, fan and dehumidification module based on the safety constraint judgment results, the natural ventilation gating results obtained in step S5, the predicted temperature rise of the hot spot inside the cabinet obtained in step S2, the current control status obtained in step S3 and the interval control parameters obtained in step S4. When the safety constraint judgment result is abnormal, a safety interlock protection command is generated, which prohibits energy-saving power reduction control and natural ventilation energy-saving control, and prioritizes the execution of forced cooling, forced dehumidification, backup air conditioning switching or alarm transmission. When the safety constraint judgment result is normal and the natural ventilation gating result is to prohibit natural ventilation, a command to prohibit natural ventilation is generated, and dehumidification control or cooling and dehumidification control is executed according to the indoor relative humidity and humidity threshold. When the safety constraint judgment result is normal, the natural ventilation gate control result is not to prohibit natural ventilation, and the predicted value of the hot spot temperature rise in the cabinet reaches the pre-cooling trigger threshold corresponding to the current control state, a predicted pre-cooling command is generated, and the air conditioning cooling power is increased or the air conditioning set temperature is reduced. When no safety interlock protection command, natural ventilation prohibition command, or pre-cooling prediction command is generated, a dehumidification control command, a natural ventilation energy-saving control command, or a conventional energy-saving cooling command is generated based on the current control status, zone control parameters, indoor relative humidity, and natural ventilation gating results. S7. Execute the control command generated in step S6, and re-collect the temperature of the hot spot node in the cabinet and the indoor ambient temperature after the preset control cycle to determine the actual temperature rise of the hot spot in the cabinet. The actual temperature rise of the hot spot inside the cabinet is compared with the predicted temperature rise of the hot spot inside the cabinet output in step S2. When the prediction deviation exceeds the preset deviation threshold, the pre-cooling trigger threshold, preset safety margin, or load rate trend-hot spot temperature rise coupled prediction model parameters corresponding to the current control state are corrected to form a closed-loop control.

2. The energy-saving control method for air conditioning in a substation switchgear room considering transformer load rate according to claim 1, characterized in that, In step S1, the hot spot nodes inside the cabinet include at least one of the following: busbar connection, circuit breaker contact, cable joint, disconnector contact, and heating connection terminal inside the cabinet. The temperature of the hot spot nodes inside the cabinet is collected by a fiber optic temperature sensor, an infrared temperature sensor, or a contact temperature sensor.

3. The energy-saving control method for air conditioning in a substation switchgear room considering transformer load rate according to claim 1, characterized in that, In step S1, the current hot spot temperature rise in the cabinet is the difference between the temperature of the hot spot node in the cabinet and the indoor ambient temperature. When there are multiple hot spot nodes in the cabinet, the maximum value of the temperature rise of each hot spot node in the cabinet is taken as the current hot spot temperature rise in the cabinet, or the risk equivalent hot spot temperature rise is calculated according to the equipment risk weight of different hot spot nodes in the cabinet, and the risk equivalent hot spot temperature rise is taken as the current hot spot temperature rise in the cabinet.

4. The energy-saving control method for air conditioning in a substation switchgear room considering transformer load rate according to claim 1, characterized in that, In step S1, the transformer load rate satisfies β=100%×S / Sn, where S is the actual apparent power of the transformer, Sn is the rated capacity of the transformer, and the unit of β is %; the load rate change rate is determined based on the difference between the transformer load rate at the current sampling time and the transformer load rate at the previous statistical time or the beginning of the sliding time window. The load rate change sequence is composed of the transformer load rates in multiple consecutive sampling periods in chronological order.

5. The energy-saving control method for air conditioning in a substation switchgear room considering transformer load rate according to claim 1, characterized in that, In step S2, the load rate trend-hot spot temperature rise coupled prediction model is a long short-term memory network model. The input features of the long short-term memory network model include at least the transformer load rate, load rate change rate, load rate change sequence, indoor ambient temperature, indoor relative humidity, top oil temperature, current hot spot temperature rise in the cabinet and air conditioning operation status. The output is the predicted value of the hot spot temperature rise in the cabinet within the next 10 to 30 minutes.

6. The energy-saving control method for air conditioning in a substation switchgear room considering transformer load rate according to claim 1, characterized in that, In step S3, the light load zone is the state where the transformer load rate is less than 30% and the load rate change rate is less than the preset surge threshold; the medium load zone is the state where the transformer load rate is greater than or equal to 30% and less than or equal to 70%; the heavy load zone is the state where the transformer load rate is greater than 70%; and the load surge zone is the state where the transformer load rate is less than or equal to 70% but the load rate change rate reaches the preset surge threshold, or the state where the transformer load rate will be greater than 70% in the future preset period based on the extrapolation of the load rate change sequence.

7. The energy-saving control method for air conditioning in a substation switchgear room considering transformer load rate according to claim 1, characterized in that, In step S4, the first pre-cooling trigger threshold is 6℃, the second pre-cooling trigger threshold is 5℃, and the third pre-cooling trigger threshold is 4℃. In the light load area, the basic control priority is dehumidification, natural ventilation, and intermittent cooling, which decreases in that order. In the medium load area, the air conditioning set temperature is corrected according to the formula Tset=32-0.05, where Tset is the transformer load rate expressed as a percentage. In the heavy load area and the load surge area, the air conditioning set temperature is lowered to 26℃ to 28℃, and the pre-cooling control is started in advance when the predicted temperature rise of the hot spot in the cabinet reaches the third pre-cooling trigger threshold.

8. The energy-saving control method for air conditioning in a substation switchgear room considering transformer load rate according to claim 1, characterized in that, In step S5, the surface temperature for judging condensation inside the cabinet is the surface temperature collected by temperature sensors located on the metal surface of the switch cabinet, the surface near the busbar support, the surface near the cable joint, or the surface near the circuit breaker contact. When multiple surface temperature collection points are set, the lowest value among the multiple surface temperatures is used as the surface temperature for judging condensation inside the cabinet.

9. The energy-saving control method for air conditioning in a substation switchgear room considering transformer load rate according to claim 1, characterized in that, In step S6, the safety constraint judgment result is determined based on the temperature of the hot spot node inside the cabinet, the indoor ambient temperature, the indoor relative humidity, the transformer load rate, the winding temperature, and the air conditioner operating status. When the temperature of the hot spot node inside the cabinet reaches the first hard constraint threshold, the indoor ambient temperature reaches the second hard constraint threshold, the indoor relative humidity reaches the third hard constraint threshold, the transformer is overloaded or overheated, or the air conditioner malfunctions, the safety constraint judgment result is abnormal and the system enters the safety interlock protection.

10. The energy-saving control method for air conditioning in a substation switchgear room considering transformer load rate according to claim 1, characterized in that, In step S6, when no safety interlock protection command, natural ventilation prohibition command, or pre-cooling prediction command is generated, fuzzy PID adjustment is performed based on the indoor temperature deviation and the rate of change of temperature deviation to dynamically adjust the PID parameters and output the air conditioner compressor frequency command and fan speed command; wherein, the indoor temperature deviation is the difference between the indoor ambient temperature and the air conditioner set temperature determined in step S4.

Images