Environment temperature control method for communication machine room of transformer substation

Abstract

The invention relates to the technical field of intelligent control, in particular to an environment temperature control method for a transformer substation communication machine room, and the method comprises the steps: collecting the environment temperature and server monitoring data in the transformer substation communication machine room; a first error optimization factor is obtained through equipment activity prediction, and preliminary error correction is carried out; acquiring a second error optimization factor through actual temperature response and prediction deviation evaluation, and performing final error correction; and obtaining optimized temperature control output through the final error correction result. And intelligent temperature regulation and control are performed through the optimized temperature control output, so that the environment temperature control precision of the communication machine room of the transformer substation is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent control technology, and particularly relates to a method for controlling the ambient temperature of a substation communication machine room. Background Art

[0002] The substation communication machine room is an indispensable key infrastructure for the stable and reliable operation of modern power systems. A large number of precision electronic devices that perform functions such as data acquisition, transmission, processing, monitoring, and protection control are centrally deployed inside it, such as servers, data storage arrays, network switches, routers, optical transmission devices, and communication interface units of various intelligent electronic devices. These communication and information processing devices have strict requirements for the operating environmental conditions, especially the ambient temperature. Excessive, too low, or drastic fluctuations in temperature will lead to a decline in device performance, unstable operation, data transmission errors, and even shorten the service life of the device or cause permanent damage. In order to ensure that the ambient temperature in the communication machine room is maintained within the appropriate working range recommended by the device manufacturer, the machine room is generally equipped with a dedicated air conditioning system, which adjusts and controls the indoor temperature by continuous or intermittent operation.

[0003] In existing technical practices, for the control of the air conditioning system in the substation communication machine room, the widely adopted control strategy is the PID (Proportional-Integral-Derivative) control strategy based on a fixed temperature set point. The core logic of this control strategy is: through temperature sensors deployed at specific positions in the machine room (usually the return air outlet or a representative area), continuously monitor the air temperature in the machine room in real time; compare the real-time monitored temperature value with a pre-set target temperature value to calculate the current temperature deviation signal; the PID controller calculates a control output signal according to the magnitude of the deviation signal, the accumulation of the deviation over time, and the change rate of the deviation, in accordance with the classical PID control algorithm; this control signal is then used to drive the actuators of the air conditioning system, such as adjusting the operating frequency of the variable frequency compressor, controlling the opening of the chilled water valve, and adjusting the rotational speed of the supply air fan, etc., to change the cooling power output of the air conditioning system. The ultimate goal is to maintain the machine room temperature near the set point. Under the condition of relatively stable or slowly changing internal heat load in the machine room, this traditional PID control method based on single temperature feedback, due to its simple structure, easy implementation and adjustment, can provide basic temperature control functions and meet the operation requirements to a certain extent.

[0004] However, the aforementioned PID environmental control method based on fixed setpoints and single-temperature feedback exposes its inherent technical limitations when dealing with the inherently dynamic internal heat load characteristics in the actual operation of a substation communication machine room, and it is difficult to fully meet the dual requirements of precise temperature control and operating energy efficiency. Specifically, the core problem faced by this technical solution is that its control response highly depends on the feedback variable of the overall air temperature in the machine room, which has significant hysteresis, and it cannot effectively cope with the rapid and local heat load shocks caused by the rapid changes in the working states of internal communication devices. Therefore, there is an inherent contradiction between suppressing the risk of transient temperature overshoot and ensuring the operating energy efficiency of the system. The details are elaborated as follows: The information processing units such as servers and network devices in the substation communication machine room often do not have constant computing tasks, data processing volumes, or communication traffic, but show significant dynamics. When performing periodic large-capacity data backups, responding to intensive information interactions caused by power grid faults, handling sudden network traffic peaks, or conducting system diagnosis and maintenance operations, the instantaneous power consumption and the heat generated by these devices will rapidly increase within a short period (in minutes or even shorter), forming pulsed or stepwise internal heat source disturbances. Due to the inevitable thermal inertia of the air itself and the heat transfer process between the equipment and the air, there is an inherent time delay between the sudden increase in the heat generation power of the equipment and the actual measurement of the temperature sensor selected as the feedback point in the machine room to obtain a temperature rise signal sufficient to drive the PID controller to make significant adjustments. This response lag characteristic of the control loop makes the traditional PID controller essentially a "post-event" response, that is, the cooling output starts to be adjusted only after the temperature has started to deviate significantly from the target, or even after a potential overheating risk has formed in the local area near the equipment. This lag response directly leads to the situation that when facing a rapid internal heat load shock, the control system often cannot timely and fully provide the required cooling capacity to "absorb" this additional heat, easily causing a short but large upward fluctuation in the controlled temperature (especially the temperature at the equipment inlet or around key components that has a more direct impact on the equipment operation), namely the "transient temperature overshoot" phenomenon. Even if this transient overshoot lasts for a short time, it will cause thermal stress shocks to highly sensitive electronic components, affecting their long-term operation stability and reliability, and even triggering the overheat protection mechanism of the equipment itself. On the contrary, if the setpoint of the PID controller is set too low or a larger control gain is adopted in an attempt to avoid this risk to speed up the response, when the equipment is operating at low load or in a steady state, it is very easy to cause unnecessary excessive cooling, resulting in the air conditioning system running at a high power state for a long time or starting and stopping frequently. This not only significantly increases the auxiliary energy consumption of the machine room, reduces the energy utilization efficiency, but also is not conducive to the maintenance and lifespan of the air conditioning equipment itself.Therefore, when the existing technology is faced with the challenge of rapid and dynamic internal heat loads in a substation communication machine room, it is difficult to achieve an ideal balance between ensuring precise and stable temperature (effectively suppressing transient overshoot) and achieving energy conservation and consumption reduction (avoiding excessive refrigeration). There are obvious technical bottlenecks, and there is an urgent need to propose an environmental control method that is more forward-looking, adaptable, and refined for improvement. Summary of the Invention

[0005] In view of this, the present invention aims to propose an environmental temperature control method for a substation communication machine room to improve the temperature control accuracy of the substation communication machine room.

[0006] To achieve the above object, the technical solution of the present invention is realized as follows:

[0007] An environmental temperature control method for a substation communication machine room, the method comprising:

[0008] Step S1: Collect the environmental temperature inside the substation communication machine room and the server monitoring data;

[0009] Step S2: Obtain a first error optimization factor through equipment activity prediction and perform preliminary error correction;

[0010] Step S3: Obtain a second error optimization factor through actual temperature response and prediction deviation evaluation and perform final error correction;

[0011] Step S4: Obtain an optimized temperature control output based on the final error correction result;

[0012] Step S5: Perform intelligent temperature regulation through the optimized temperature control output.

[0013] Further, according to the collection of the environmental temperature inside the substation communication machine room and the server monitoring data, it specifically includes:

[0014] Obtain a set temperature sampling period, and collect temperature data through multiple temperature sensors deployed at the air return outlet of the air conditioner, hot spots, and equipment-intensive areas inside the substation communication machine room according to the set temperature sampling period to obtain temperature monitoring data;

[0015] Collect the processor utilization rate, disk activity rate, and total network interface throughput data of the servers, core switches, and routers in the substation communication machine room through the set temperature sampling period to obtain the original equipment performance index data; normalize the original equipment performance index data to obtain the normalized equipment performance index; obtain the set equipment performance index weight, and perform weighted fusion on the normalized equipment performance index through the equipment performance index weight and perform normalization processing to obtain the equipment aggregation activity level index. The calculation formula of the equipment aggregation activity level index is:

[0016] ;

[0017] Among them, represents the equipment aggregation activity level index at the time of the substation communication machine room; represents the normalization function; respectively represent the weights of three equipment performance indicators, namely the processor utilization rate, disk activity rate, and network interface throughput of the equipment; represents the th normalized index of the processor utilization rate of the th server at the time; represents the th normalized index of the disk activity rate of the th server at the

[0018] Obtain the set control parameters through equipment requirements, including the basic temperature set point, warning temperature threshold, and maximum safety temperature.

[0019] Furthermore, obtain the first error optimization factor according to the prediction of equipment activities and perform preliminary error correction, specifically including:

[0020] Obtain the equipment aggregation activity level index; take the first derivative of the equipment aggregation activity level index as the change rate of the equipment aggregation activity level index; take the second derivative of the equipment aggregation activity level index as the change acceleration of the equipment aggregation activity level index; evaluate the activity intensity through the change rate and change acceleration of the equipment aggregation activity level index to obtain the dynamic activity intensity factor; perform a non-linear mapping on the dynamic activity intensity factor to obtain the preliminary error optimization factor and perform preliminary correction on the original error through the preliminary error optimization factor to obtain the preliminarily corrected error signal.

[0021] Furthermore, according to the evaluation of the activity intensity through the change rate and change acceleration of the equipment aggregation activity level index to obtain the dynamic activity intensity factor, specifically including:

[0022] Obtain the set dynamic characteristic sensitivity time constant and acceleration term relative importance factor; the calculation formula of the dynamic activity intensity factor is:

[0023] ;

[0024] Among them, represents the dynamic activity intensity factor of the equipment in the substation communication machine room at the moment; represents the equipment aggregation activity level index of the substation communication machine room at the moment; represents the set dynamic characteristic sensitivity time constant; represents the change rate of the equipment aggregation activity level index at the moment; represents the change acceleration of the equipment aggregation activity level index at the moment; represents the modified linear function, i.e., .

[0025] Furthermore, through non-linearly mapping the dynamic activity intensity factor, a preliminary error optimization factor is obtained, and the original error is preliminarily corrected by the preliminary error optimization factor to obtain a preliminarily corrected error signal, which specifically includes:

[0026] The calculation formula of the preliminary error optimization factor is:

[0027] ;

[0028] wherein, represents the preliminary error optimization factor of the substation communication machine room at the moment; represents the set maximum prediction bias amplitude, i.e., the maximum value of the preliminary error optimization factor; represents the dynamic activity intensity factor of the equipment in the substation communication machine room at the moment; represents the reference dynamic activity intensity, i.e., the mean value of the dynamic activity intensity factors with a historical window length of 24 hours; represents the scale factor of the dynamic activity intensity, i.e., the standard deviation of the dynamic activity intensity factors with a historical window length of 24 hours; represents the hyperbolic tangent mapping function;

[0029] Obtain the monitored temperature of the substation communication machine room, and use the difference between the monitored temperature of the substation communication machine room and the set value of the base temperature as the original temperature error; use the calculation result of adding the preliminary error optimization factor and the original temperature error as the preliminarily corrected error signal.

[0030] Furthermore, through evaluating the actual temperature response and the prediction deviation, a second error optimization factor is obtained and the final error correction is performed, which specifically includes:

[0031] Obtain temperature monitoring data, the set maximum correction amplitude and the upper limit of the temperature change rate; through the change evaluation of the temperature monitoring data of the substation communication machine room, obtain the actual temperature change rate and the degree of the actual temperature approaching the upper limit; through the relevant analysis of the recent equipment activity status of the actual temperature change rate, obtain the correlation flag; through the normal fluctuation deviation evaluation of the actual temperature change rate, obtain the fluctuation deviation of the actual temperature; through the fusion evaluation of the degree of the actual temperature approaching the upper limit, the correlation flag and the fluctuation deviation, obtain the second error optimization factor, and use the second error optimization factor to perform the final error correction on the preliminarily corrected error signal to obtain the final error signal.

[0032] Further, according to the change evaluation of the temperature monitoring data of the substation communication machine room, obtain the actual temperature change rate and the degree of the actual temperature approaching the upper limit; through the relevant analysis of the recent equipment activity status of the actual temperature change rate, obtain the correlation flag; through the normal fluctuation deviation evaluation of the actual temperature change rate, obtain the fluctuation deviation of the actual temperature, specifically including:

[0033] Obtain the set temperature sampling period, and use the calculation result of dividing the difference between two consecutive temperature sampling monitoring values by the temperature sampling period as the actual temperature change rate;

[0034] Obtain the warning temperature threshold for triggering the correction mechanism and the maximum allowable safe temperature of the substation communication machine room. The calculation formula for the degree of the actual temperature approaching the upper limit is:

[0035] ;

[0036] Among them, represents the degree of the actual temperature of the substation communication machine room approaching the upper limit at moment; represents the actual monitored temperature of the substation communication machine room at moment; represents the warning temperature threshold for triggering the correction mechanism; represents the maximum allowable safe temperature of the substation communication machine room;

[0037] Obtain the set time window for associated flag evaluation, rapid temperature increase threshold, high activity intensity threshold, rapid change rate threshold, acceleration threshold, and logical operation dual judgment conditions. The evaluation process of the associated flag is to perform logical judgment through the logical operation function of the logical operation dual judgment conditions. Condition 1 in the logical operation dual judgment conditions is that the current actual temperature change rate exceeds the set rapid temperature increase threshold; Condition 2 in the logical operation dual judgment conditions is whether there is a situation of high device activity or violent dynamic change within the set time window for associated flag evaluation. The judgment conditions for the situation of high device activity or violent dynamic change include whether there is any moment within the time window for associated flag evaluation that satisfies any of the following conditions:

[0038] 1. The device aggregated activity level index is greater than the high activity intensity threshold;

[0039] 2. The change rate of the device aggregated activity level index is greater than the rapid change rate threshold;

[0040] 3. The change acceleration of the device aggregated activity level index is greater than the acceleration threshold;

[0041] When and only when both Condition 1 and Condition 2 are satisfied, the value of the associated flag is , otherwise the value is ;

[0042] Obtain the upper limit of the acceptable temperature change rate, and use the calculation result of subtracting the actual temperature change rate from the upper limit of the acceptable temperature change rate as the fluctuation deviation of the actual temperature.

[0043] Furthermore, through the fusion evaluation of the degree of actual temperature approaching the upper limit, the associated flag, and the fluctuation deviation, obtain the second error optimization factor, and perform final error correction on the preliminarily corrected error signal through the second error optimization factor to obtain the final error signal. Specifically, it includes:

[0044] Obtain the set maximum value of the second error optimization factor. The calculation formula of the second error optimization factor is:

[0045] ;

[0046] Among them, represents the second error optimization factor of the substation communication machine room at moment; represents the set maximum value of the second error optimization factor; represents the associated flag of the substation communication machine room at moment; represents the degree of actual temperature approaching the upper limit of the substation communication machine room at moment; Indicates the fluctuation deviation of the actual temperature of the substation communication machine room at the moment;

[0047] Obtain the preliminarily corrected error signal, and use the calculation result of adding the preliminarily corrected error signal and the second error optimization factor as the final error signal.

[0048] Furthermore, according to the optimized temperature control output obtained from the final error correction result, specifically including:

[0049] Use the final error signal as the input and apply it to a standard PID controller algorithm to perform proportional, integral, and differential operations on the input final error signal, obtain the adjustment control signal output by the PID controller algorithm, and use this adjustment control signal as the optimized temperature control output.

[0050] Compared with the prior art, the present invention has the following advantages:

[0051] An environmental temperature control method for a substation communication machine room according to the present invention uses device activity information for forward-looking prediction and combines feedback on the actual temperature response for dynamic correction to generate a highly optimized error signal to drive the controller. This mechanism enables the control system to anticipate and respond in advance to potential thermal load shocks caused by rapid changes in internal device activities, significantly shortening the control response time, and thus effectively suppressing the transient temperature overshoot phenomenon that is prone to occur under traditional control. At the same time, due to the accuracy of compensation and the existence of the correction mechanism, unnecessary overcooling can be avoided when the device load is stable or decreasing. Therefore, the method of the present invention can ultimately effectively respond to the challenge of internal dynamic thermal load on the premise of ensuring that the temperature of the communication machine room is accurately and stably maintained within the safe operating range, improve the operating energy efficiency of the air conditioning system, and successfully solve the technical problem of the difficult balance between temperature stability and energy conservation and consumption reduction described in the background art. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The drawings constituting a part of the present invention are used to provide a further understanding of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0053] Figure 1 is a method flow chart of an environmental temperature control method for a substation communication machine room according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0054] It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other.

[0055] In the description of the present invention, it should be noted that the orientation or positional relationship indicated by the terms "upper", "lower", "inner", "back", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present invention. In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance.

[0056] The present invention will be described in detail below with reference to the drawings and in conjunction with embodiments.

[0057] See Figure 1 , which is a method flow chart of an environmental temperature control method for a substation communication machine room provided in the first embodiment of the present invention. As Figure 1 shown, an environmental temperature control method for a substation communication machine room may include:

[0058] S1, collect the environmental temperature inside the substation communication machine room and the server monitoring data.

[0059] Obtain the set temperature sampling period, and collect temperature data according to the set temperature sampling period through multiple temperature sensors deployed at the air return outlet of the air conditioner, hot spots, and equipment-intensive areas inside the substation communication machine room to obtain temperature monitoring data;

[0060] Collect the processor utilization rate, disk activity rate, and total network interface throughput data of the servers, core switches, and routers in the substation communication machine room through the set temperature sampling period to obtain the original device performance index data; normalize the original device performance index data to obtain the device performance normalization index; obtain the set device performance index weight, and perform weighted fusion on the device performance normalization index through the device performance index weight and perform normalization processing to obtain the device aggregation activity level index. The calculation formula of the device aggregation activity level index is:

[0061]

[0062] Wherein, represents the device aggregation activity level index of the substation communication machine room at time; represents the normalization function; respectively represent the weights of the three device performance indicators of the processor utilization rate, disk activity rate, and network interface throughput of the device; represents the th server at time of the normalization index of the processor utilization rate; Indicates the normalized index of the disk activity rate of the th server at moment; Indicates the normalized index of the throughput of the th network interface at moment; Indicates the number of server devices; Indicates the number of network interfaces.

[0063] Obtain the set control parameters through device requirements, including the basic temperature set point, warning temperature threshold and maximum safe temperature.

[0064] S2, obtain the first error optimization factor through device activity prediction and perform preliminary error correction.

[0065] The core technical problem to be solved by the present invention lies in the inherent limitation of the traditional environmental control method in information acquisition. Specifically, this method completely relies on the real-time monitoring result of the physical quantity of the air temperature in the computer room as its only feedback input. However, the air temperature itself is a variable with significant thermal inertia. Inside the substation communication computer room, the main source of heat is the operating electronic devices. When the working states of these devices (such as servers performing intensive computing tasks and network devices processing burst traffic) change, especially when they rapidly transition from a low load to a high load state, their internal power consumption and corresponding heat generation power will increase instantaneously. But since the heat needs to be first transferred to the surrounding air through the device's own heat dissipation structure, and then gradually affect the readings of the temperature sensors deployed at specific locations (such as the air conditioner return air outlet) through the convection and mixing processes of the air, this entire physical transfer chain is inevitably accompanied by a time delay. Therefore, there is an inevitable lag between the temperature change sensed by the temperature sensor and the actual change in the device's heat generation power.

[0066] Traditional controllers, due to their design principle being based on the response to the deviation between the current measured temperature and the set temperature, this lag of the input signal is directly converted into a lag of the control action. In other words, the controller can only start adjusting the cooling output of the air conditioning system after the temperature has deviated significantly. In scenarios where the device heat load is stable or changes slowly, this lag can be tolerated. However, for the operating states of the devices in the substation communication computer room, according to the actual business requirements (such as periodic data backup, centralized reporting of power grid fault information, remote diagnosis operations, etc.), if there are sudden, pulsed or stepwise high-intensity activities, it will cause its heat generation power to increase rapidly and significantly in a short period of time. In this context of rapid dynamic thermal disturbances, traditional The hysteretic response characteristic of the controller then becomes a serious defect. During this "response time" when it has not yet been able to effectively increase the refrigeration power to cope with the newly added heat load, the heat quickly released by the equipment will rapidly accumulate in the surrounding space, especially in the key areas affecting the equipment's heat dissipation efficiency (such as near the air inlet of the equipment), very easily leading to a transient sharp increase in the local temperature, forming the so-called "transient temperature overshoot". Such a temperature overshoot, even if it lasts for a short time, will also pose a threat to the high-precision and high-sensitivity communication electronic equipment in the computer room, affecting its operation stability and long-term reliability.

[0067] In order to overcome the above-mentioned problem of insufficient foresight caused by relying on hysteretic temperature feedback, the present invention proposes a preliminary optimization mechanism based on the prediction of equipment activity status in step S2. Instead of passively waiting for the change of temperature signal, it actively captures the pre-information that can predict the future change of heat load. Given that the main source of heat inside the computer room is the operating power consumption of electronic equipment, then directly monitoring the working activity levels of these key equipment (by collecting parameters such as the utilization rate of the server, hard disk , network interface throughput, etc. that can reflect its computing or communication intensity, and comprehensively evaluating to obtain an aggregated activity level index ), it is possible to obtain information that can reflect the change of the internal heat source intensity earlier and more directly than the air temperature.

[0068] Merely obtaining the current equipment activity level is not sufficient to fully achieve forward-looking prediction. Because the impact intensity of the heat load is not only related to the current activity level, but also closely related to its dynamic characteristics of change. A state of continuously high activity level and a state that rapidly jumps from a low level to a high level, the latter poses an obviously greater challenge to the temperature control system. Therefore, this step further focuses on the change rate and the acceleration of change . Among them, the change rate reflects the current increasing or decreasing trend of the activity level, while the acceleration reveals the intensity of this trend and is the key feature to judge whether there is a "sudden" change. A significantly positive acceleration signal strongly indicates that the equipment activity is increasing sharply, and the heat load will probably rapidly climb in the near future.

[0069] Based on the above analysis, the goal of step S2 is to utilize the real-time obtained and calculated aggregated activity level of the equipment and its dynamic characteristics and Quantitatively evaluate the expected heat load intensity that is indicated by the current device activity status and will affect the temperature of the computer room in the near future. Then, convert this quantified expected intensity into an equivalent temperature offset value . This is not the temperature deviation measured actually, but a predictive and additional error signal, which aims to simulate the influence of the upcoming temperature change trend on the controller

[0070] In summary, obtain the device aggregation activity level index; take the first derivative of the device aggregation activity level index as the change rate of the device aggregation activity level index; take the second derivative of the device aggregation activity level index as the change acceleration of the device aggregation activity level index; evaluate the activity intensity through the change rate and change acceleration of the device aggregation activity level index to obtain the dynamic activity intensity factor; perform a non-linear mapping on the dynamic activity intensity factor to obtain the preliminary error optimization factor and preliminarily correct the original error through the preliminary error optimization factor to obtain the preliminarily corrected error signal

[0071] Obtain the set dynamic characteristic sensitivity time constant and the relative importance factor of the acceleration term; the calculation formula of the dynamic activity intensity factor is

[0072] ;

[0073] wherein represents the dynamic activity intensity factor of the device in the substation communication computer room at time; represents the device aggregation activity level index of the substation communication computer room at time; represents the set dynamic characteristic sensitivity time constant, and set the dynamic characteristic sensitivity time constant according to the heat response time of the computer room Set the dynamic characteristic sensitivity time constant. For the estimated heat response time of a typical medium and small substation communication computer room, it is minutes. In the embodiment of the present invention, the set dynamic characteristic sensitivity time constant is times of i.e., ; represents the change rate of the device aggregation activity level index at time; represents the relative importance factor of the acceleration term. In the embodiment of the present invention, the set initial value of is represents the change acceleration of the device aggregation activity level index at time; represents the correction linear function, i.e.,

[0074] It should be noted that this design is not a simple linear weighting, but rather a structured non-linear combination. The basic activity level is used as a basis and multiplied by an enhancement factor determined by dynamic features ( ). This structure enables the effects of dynamic changes (especially activity increase and accelerating increase , through function screening) to be combined in the form of the square root of the sum of squares, simulating the "dynamic shock amplitude" under the combined action of the rate of change and acceleration. Its non-linear characteristics can better reflect the intensity of the combined effect. At the same time, the overall sensitivity of the dynamic part is controlled through parameter , and the importance of acceleration , which can better reflect the "suddenness" characteristic, is adjusted and allowed to be highlighted through parameter , directly addressing the core problem that the existing technology has difficulty coping with rapid and pulsed heat loads.

[0075] After obtaining the dynamic activity intensity factor, a non-linear mapping can be performed on the dynamic activity intensity factor to obtain a preliminary error optimization factor, and the original error can be preliminarily corrected through the preliminary error optimization factor to obtain a preliminarily corrected error signal. The calculation formula for the preliminary error optimization factor is:

[0076] ;

[0077] where represents the preliminary error optimization factor of the substation communication machine room at time; represents the set maximum prediction bias amplitude, i.e., the maximum value of the preliminary error optimization factor; represents the dynamic activity intensity factor of the equipment in the substation communication machine room at time; represents the reference dynamic activity intensity, i.e., the mean value of the dynamic activity intensity factors with a historical window length of 24 hours; represents the scale factor of the dynamic activity intensity, i.e., the standard deviation of the dynamic activity intensity factors with a historical window length of 24 hours; represents the hyperbolic tangent mapping function.

[0078] It should be noted that the hyperbolic tangent function is used to map the calculated dynamic activity intensity factor to an error bias value with actual physical meaning (temperature unit). The application of the function ensures the smoothness of the mapping process, avoids sudden changes in the control signal, and is beneficial to system stability; at the same time, its inherent saturation characteristic will The output limit is within a preset maximum amplitude , which not only conforms to the finiteness of prediction compensation in physical reality but also enhances the robustness of the algorithm in the face of extreme inputs. Introduce a reference point and a scale factor such that the center and sensitivity of the mapping can be set according to the statistical characteristics of . Finally, the calculated based on device activity prediction is superimposed on the original temperature error to form a corrected error signal . When device activity indicates that the future heat load will increase (e.g., ), will be positive, causing , thus driving the

[0079] controller to take cooling actions in advance, effectively shortening the response time to internal thermal disturbances.

[0080] During specific implementation, the calculated preliminary error bias factor is superimposed on the original error signal calculated from the actual temperature and the basic set point to form a preliminarily optimized error signal . When the monitored device activity characteristics (high activity level, rapid increase rate, significant positive acceleration) indicate that the future heat load will increase significantly, will be calculated as a positive value. At this time, even if the actual temperature has not increased significantly, the corrected error signal will be larger than the original error . This "artificially" amplified error signal is input into the controller, which will drive the controller to react in advance, that is, to command the air conditioning system to increase the cooling output earlier.

[0081] S3. Obtain a second error optimization factor through the evaluation of the actual temperature response and prediction deviation and perform final error correction.

[0082] Step S2 introduces a preliminary error bias factor based on device activity prediction , aiming to utilize device activity information that responds faster than temperature, adjust the error input of the PID controller in advance, and thus achieve a proactive response to internal heat load shocks. However, even under ideal conditions (i.e., the device activity information is accurate and its basic relationship with heat generation meets expectations), there are still inherent limitations, which constitute the direct and necessary reasons for introducing step S3 for optimization:

[0083] Quantization accuracy limitation of the prediction model: Step S2 constructs a mapping relationship from device activity characteristics to predictive temperature offsets . Essentially, it is a quantitative estimation of a complex thermodynamic process. Although the setting of parameters ( ) has its physical and statistical basis, they represent a relationship under an average or typical situation. Under specific operating conditions, the actual heat effect generated by the device activity state (such as a specific pattern), as well as the exact time required for this heat effect to be transmitted to the temperature sensor and the resulting exact temperature rise amplitude, will be affected by multiple complex and dynamically changing factors such as the specific airflow distribution inside the computer room at present, the exact position of the device, and the instantaneous operating efficiency of the air conditioning system. Therefore, the value calculated from the fixed model structure and parameters, although correct in direction (indicating a temperature increase or decrease) and trend (reflecting the severity), there is an inevitable deviation in the specific magnitude from the equivalent temperature rise that actually needs to be compensated at that moment. That is, may be quantitatively underestimated at some moments, or overestimated at other moments.

[0084] Time matching deviation between prediction and actual response: The goal of step S2 is to make act in advance to compensate for the thermal lag of the physical system. However, the actual heat transfer and temperature response processes are extremely complex, and its equivalent lag time itself is not a constant value, but will vary slightly with factors such as airflow and load distribution. The calculated based on a model with fixed parameters and a fixed structure, its acting timing cannot perfectly match the advance amount actually required in all cases.

[0085] Due to the inherent limitations of the above prediction model in quantitative accuracy and timing matching, simply relying on the preliminary optimization of step S2 cannot achieve the ideal control effect in all cases. Especially when the device activity changes extremely violently or the pattern is relatively special, the prediction deviation of

[0086] will cause the control system to still be unable to completely avoid transient temperature overshoots, or generate unnecessary control fluctuations. Therefore, the core purpose of step S3 is to establish a real-time evaluation and correction mechanism based on the actual system response, specifically for monitoring and compensating step The possible biases that the prediction mechanism may generate. This step does not focus on those minor or external interference sources, but directly focuses on whether the actual performance of the controlled temperature after applying the prediction compensation meets the expectations (i.e., the temperature stabilizes near the set point , or at least changes smoothly).

[0087] The specific logic is as follows: Step S3 continuously monitors the actual computer room temperature and its change rate . When it is found that still shows an abnormally rapid upward trend ( significantly exceeding the normal range ), and this abnormal rise occurs at a critical moment when the temperature has approached the warning line or (quantified by ), and at the same time, through correlation analysis it is confirmed that this abnormal temperature rise does coincide in time with the current or recent high activity state of the equipment (i.e., the core disturbance source that Step S2 attempts to address), the system then judges that: the current prediction bias generated by Step S2 is quantitatively insufficient under the current specific working conditions and fails to effectively suppress the temperature rise driven by equipment activities.

[0088] In this scenario of rapid temperature rise related to equipment activities and endangering the temperature upper limit, which is proven to be caused by insufficient prediction (relative to actual needs), Step S3 activates a correction mechanism. Calculate a corrective error bias factor , the magnitude of which is directly related to the severity of the observed prediction deviation (i.e., the degree of exceeding the normal range and the degree of approaching the upper limit ). This is not for a new disturbance source, but a real-time calibration and supplement to itself. Add to to obtain the final error signal , which is equivalent to dynamically and on-demand adding an additional error bias based on the actually observed control effect deviation on the basis of , driving the controller to generate a stronger immediate response to make up for the quantitative insufficiency at that specific moment, thereby more effectively suppressing the risk of critical temperature overshoot.

[0089] In summary, the specific process of this step is to obtain temperature monitoring data, set the maximum correction amplitude and the upper limit of the temperature change rate; by evaluating the change of the temperature monitoring data in the substation communication room, obtain the actual temperature change rate and the degree of the actual temperature approaching the upper limit; by analyzing the actual temperature change rate in relation to the recent equipment activity status, obtain an association flag; by evaluating the normal fluctuation deviation of the actual temperature change rate, obtain the fluctuation deviation of the actual temperature; by comprehensively evaluating the degree of the actual temperature approaching the upper limit, the association flag and the fluctuation deviation, obtain a second error optimization factor, and use the second error optimization factor to perform final error correction on the preliminarily corrected error signal to obtain the final error signal.

[0090] According to the specific process of this step, first, obtain the set temperature sampling period, and use the calculation result of dividing the difference between two consecutive temperature sampling monitoring values by the temperature sampling period as the actual temperature change rate;

[0091] Obtain the warning temperature threshold for triggering the correction mechanism and the maximum allowable safe temperature of the substation communication room. The formula for calculating the degree of the actual temperature approaching the upper limit is:

[0092] ;

[0093] Wherein, represents the degree of the actual temperature approaching the upper limit in the substation communication room at time; represents the actual monitored temperature in the substation communication room at time; represents the warning temperature threshold for triggering the correction mechanism, which is set to in the embodiment of the present invention; represents the maximum allowable safe temperature of the substation communication room.

[0094] Next, determine whether the current temperature change is significantly related to the recent equipment activity status to obtain an association flag :

[0095]

[0096] Wherein, represents the association flag in the substation communication room at time; represents the logic operation function of the double judgment condition of the logic operation; represents the actual temperature change rate of the substation communication room at time; is the logic operation parameter.

[0097] It should be noted that the evaluation process of the associated mark is to obtain the set time window for the associated mark evaluation, the rapid temperature rise threshold, the high activity intensity threshold, the rapid change rate threshold, the acceleration threshold and the logical operation dual judgment condition. The evaluation process of the associated mark is to perform logical judgment through the logical operation function of the logical operation dual judgment condition. The first condition of the logical operation dual judgment condition is that the current actual temperature change rate exceeds the set rapid temperature rise threshold; the second condition of the logical operation dual judgment condition is whether there is a situation of high device activity or drastic dynamic changes in the time window set for the associated mark evaluation. The judgment condition for the situation of high device activity or drastic dynamic changes includes whether any of the following conditions are met at any time in the time window of the associated mark evaluation:

[0098] 1. The device aggregate activity level indicator is greater than the high activity intensity threshold;

[0099] 2. The rate of change of the device aggregate activity level indicator is greater than the rapid change rate threshold;

[0100] 3. The acceleration of the change of the device aggregate activity level indicator is greater than the acceleration threshold;

[0101] If and only if both condition 1 and condition 2 are met, the association flag takes the value , otherwise the value is .

[0102] In the embodiment of the present invention, the time window length for evaluating the association flag is set to , that is, each window includes temperature sampling points; set the rapid temperature rise threshold to (This value is based on the upper limit of normal temperature change rate. times); set the high activity intensity threshold to , which represents the system reaching its maximum activity capacity The rapid change rate threshold is set to , which represents the rate at which the activity level increases significantly in a short period of time; the acceleration threshold is set to , which indicates that the activity growth trend is accelerating sharply.

[0103] After obtaining the associated mark of the substation communication room, the normal fluctuation deviation of the actual temperature change rate can be evaluated to obtain the fluctuation deviation of the actual temperature. Specifically, the acceptable upper limit of the temperature change rate is obtained, and the calculation result of subtracting the actual temperature change rate from the acceptable upper limit of the temperature change rate is used as the fluctuation deviation of the actual temperature. In the embodiment of the present invention, the acceptable upper limit of the temperature change rate of the substation communication room is set to .

[0104] Finally, by fusing and evaluating the degree of proximity of the actual temperature to the upper limit, the correlation flag, and the fluctuation deviation, a second error optimization factor is obtained, and the initially corrected error signal is finally error-corrected by the second error optimization factor to obtain the final error signal.

[0105] Specifically, the maximum value of the set second error optimization factor is obtained, and the calculation formula of the second error optimization factor is:

[0106] ;

[0107] where represents the second error optimization factor of the substation communication machine room at time; represents the maximum value of the set second error optimization factor. In the embodiment of the present invention, is set; represents the correlation flag of the substation communication machine room at time; represents the degree of proximity of the actual temperature to the upper limit of the substation communication machine room at time; represents the fluctuation deviation of the actual temperature of the substation communication machine room at time;

[0108] It should be noted that the core purpose of constructing the correction error bias factor in step S3 of the present invention is to provide a real-time evaluation and compensation mechanism based on the actual system response for the inherent quantitative accuracy and timing matching limitations of the prediction mechanism in step S2 itself. The generation of the second error optimization factor is gated and modulated by three key conditions: , , and , ensuring that it only intervenes at the most needed and relevant moments and corrects with an appropriate intensity.

[0109] First of all, the correlation flag plays a key "cause confirmation" role. It ensures that the trigger of must be associated with the core problem to be solved by the present invention, that is, the heat load impact caused by internal device activities. Only when the monitored abnormally rapid temperature rise ( ) is confirmed to coincide with recent high device activities or drastic changes in time ( ), may be activated. This effectively avoids unnecessary over-reactions of the control system to temperature fluctuations caused by other irrelevant factors (such as external disturbances), ensuring the pertinence and accuracy of the correction mechanism.

[0110] Secondly, the temperature proximity factor introduces a risk awareness. It makes the correction intensity proportional to the degree to which the current temperature approaches the upper limit of danger When the temperature is far below the warning threshold , it is , and even if there is a temperature rise caused by a prediction deviation, as long as the risk is not high it will not be activated, avoiding over-intervention in the safe area. When the temperature enters into the warning interval it linearly increases from to , making the correction effect gradually strengthen as the risk increases. This design conforms to the principle of safety first, ensuring that the correction force is concentrated on preventing the most critical over-temperature risk.

[0111] Furthermore, the core driving term quantifies the response to the degree of "insufficient prediction". The fluctuation deviation of the actual temperature directly measures the degree to which the actual temperature change rate exceeds the normal range. The function ensures that this term is positive only when (i.e., the temperature rises too fast), triggering a positive correction bias . The function then performs a smooth and bounded non-linear mapping of the deviation degree, converting it into a contribution to the magnitude. This means that the more the actual temperature rise rate exceeds the normal range the larger the correction amount of (but saturated within the limit of ), thus more powerfully compensating for the

[0112] After obtaining the second error optimization factor, obtain the preliminarily corrected error signal, and use the calculation result of adding the preliminarily corrected error signal and the second error optimization factor as the final error signal.

[0113] So far, the second error optimization factor is obtained through the actual temperature response and prediction deviation evaluation and the final error correction is performed.

[0114] S4. Obtain the optimized temperature control output through the final error correction result.

[0115] After the optimized calculations in steps S2 and S3, an optimized error signal that can finally reflect the current actual temperature deviation, predictive heat load change, and real-time correction of prediction deviation is obtained After that, this step will utilize this optimized error signal to calculate the control output command of the air conditioning system.

[0116] Specifically, the final optimized error signal is used as the input and applied to a standard (Proportional-Integral-Derivative) controller algorithm. This controller, based on its pre-set proportional gain , integral gain and derivative gain parameters, performs proportional, integral and derivative operations on the input optimized error (i.e., follows the calculation logic of , but note that the input here is the optimized instead of the original ), thereby generating a regulating control signal . This control signal represents the level or change in the air conditioning cooling power required to be adjusted according to the current (after optimized judgment) state. Here, the control algorithm itself follows the conventional implementation methods well-known in the art. The core innovation of the present invention lies in providing the optimized error input signal , rather than changing the structure of the control algorithm itself. The algorithm belongs to the existing algorithms and will not be described in too much detail here.

[0117] S5. Perform intelligent temperature regulation through the optimized temperature control output.

[0118] Convert the regulating control signal calculated in step S4 into specific operation instructions and send them to the actuators of the air conditioning system in the substation communication machine room. These actuators include the compressor frequency controller of the variable-frequency air conditioner, the electric regulating valve of the chilled water system, the speed governor of the air supply fan, etc., depending on the type of the configured air conditioning system. The control signal will drive these actuators to act and precisely adjust the actual cooling power output of the air conditioning system.

[0119] The environmental control method described in the present invention can achieve its core goal: by using the device activity information for forward-looking prediction ( ), and combining the feedback of the actual temperature response for dynamic correction ( ), generating a highly optimized error signal to drive Controller. This mechanism enables the control system to anticipate and respond in advance to potential thermal load shocks caused by rapid changes in internal device activities, significantly shortening the control response time and effectively suppressing transient temperature overshoots that are prone to occur under traditional control. At the same time, due to the accuracy of compensation and the existence of a correction mechanism, unnecessary overcooling can be avoided when the device load is stable or decreasing. Therefore, the method of the present invention can ultimately effectively address the challenges of internal dynamic thermal loads on the premise of ensuring that the temperature in the communication computer room is accurately and stably maintained within the safe operating range, improve the operating energy efficiency of the air conditioning system, and successfully solve the technical problem of the difficult balance between temperature stability and energy conservation and consumption reduction described in the background art.

[0120] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. An environmental temperature control method for a substation communication machine room, characterized in that, The method includes: Step S1: Collect the environmental temperature inside the substation communication machine room and the server monitoring data; Step S2: Obtain the first error optimization factor through device activity prediction and perform preliminary error correction; Step S3: Obtain the second error optimization factor through actual temperature response and prediction deviation evaluation and perform final error correction; Step S4: Obtain the optimized temperature control output based on the final error correction result; Step S5: Perform intelligent temperature regulation through the optimized temperature control output.

2. The environmental temperature control method for a substation communication machine room according to claim 1, wherein, According to the collection of the environmental temperature inside the substation communication machine room and the server monitoring data, it specifically includes: Obtain the set temperature sampling period, and collect temperature data through multiple temperature sensors deployed at the air return opening of the air conditioner, hot spots, and equipment-intensive areas inside the substation communication machine room according to the set temperature sampling period to obtain temperature monitoring data; Collect the processor utilization rate, disk activity rate, and total network interface throughput data of the servers, core switches, and routers in the substation communication machine room through the set temperature sampling period to obtain the original device performance index data; normalize the original device performance index data to obtain the normalized device performance index; obtain the set device performance index weight, and perform weighted fusion on the normalized device performance index through the device performance index weight and perform normalization processing to obtain the device aggregation activity level index. The calculation formula of the device aggregation activity level index is: ; Among them, represents the device aggregation activity level index of the substation communication machine room at time represents the normalization function; respectively represent the weights of three device performance indicators, namely the processor utilization rate, disk activity rate, and network interface throughput of the device; represents the th server's normalized index of processor utilization rate at time represents the th server's normalized index of disk activity rate at time represents the th network interface's normalized index of throughput at time represents the number of server devices; represents the number of network interfaces; Obtain the set control parameters through device requirements, including the basic temperature set point, warning temperature threshold, and maximum safe temperature.

3. A method for controlling the ambient temperature of a communication machine room in a substation according to claim 1, characterized in that, According to the obtaining of the first error optimization factor through device activity prediction and performing preliminary error correction, it specifically includes: Obtain the device aggregation activity level index; use the first derivative of the device aggregation activity level index as the change rate of the device aggregation activity level index; use the second derivative of the device aggregation activity level index as the change acceleration of the device aggregation activity level index; evaluate the activity intensity through the change rate and change acceleration of the device aggregation activity level index to obtain the dynamic activity intensity factor; perform non-linear mapping on the dynamic activity intensity factor to obtain the preliminary error optimization factor and perform preliminary correction on the original error through the preliminary error optimization factor to obtain the preliminarily corrected error signal.

4. A method for controlling the ambient temperature of a substation communication machine room according to claim 3, characterized in that, According to the evaluation of the activity intensity through the change rate and change acceleration of the device aggregation activity level index to obtain the dynamic activity intensity factor, it specifically includes: Obtain the set dynamic characteristic sensitivity time constant and acceleration term relative importance factor; the calculation formula of the dynamic activity intensity factor is: ； Among them, represents the dynamic activity intensity factor of the equipment in the substation communication machine room at moment; represents the equipment aggregation activity level index of the substation communication machine room at moment; represents the set dynamic characteristic sensitivity time constant; represents the change rate of the equipment aggregation activity level index at moment; represents the relative importance factor of the acceleration term; represents the change acceleration of the equipment aggregation activity level index at .

5. A method for controlling the ambient temperature of a substation communication machine room according to claim 3, characterized in that, According to the non-linear mapping of the dynamic activity intensity factor to obtain the preliminary error optimization factor and perform preliminary correction on the original error through the preliminary error optimization factor to obtain the preliminarily corrected error signal, it specifically includes: The calculation formula of the preliminary error optimization factor is: ； Among them, represents the preliminary error optimization factor of the substation communication machine room at moment; represents the set maximum prediction bias amplitude, that is, the maximum value of the preliminary error optimization factor; represents the dynamic activity intensity factor of the equipment in the substation communication machine room at moment; represents the reference dynamic activity intensity, that is, the mean value of the dynamic activity intensity factors with a historical window length of 24 hours; represents the scale factor of the dynamic activity intensity, that is, the standard deviation of the dynamic activity intensity factors with a historical window length of 24 hours; represents the hyperbolic tangent mapping function; The monitored temperature of the substation communication room is obtained, and the difference between the monitored temperature of the substation communication room and the basic temperature setting value is used as the original temperature error; the calculation result of adding the preliminary error optimization factor to the original temperature error is used as the preliminary corrected error signal.

6. A method for controlling the ambient temperature of a communication machine room in a substation according to claim 1, characterized in that, According to the method of obtaining the second error optimization factor through the actual temperature response and the predicted deviation evaluation and performing the final error correction, specifically including: Acquire temperature monitoring data, set the maximum correction amplitude and the upper limit of the temperature change rate; evaluate the changes in the temperature monitoring data of the substation communication room to obtain the actual temperature change rate and the degree to which the actual temperature approaches the upper limit; perform correlation analysis on the actual temperature change rate for recent equipment activity status to obtain an associated mark; perform a normal fluctuation deviation evaluation on the actual temperature change rate to obtain the fluctuation deviation of the actual temperature; obtain a second error optimization factor by fusion evaluation of the degree to which the actual temperature approaches the upper limit, the associated mark and the fluctuation deviation, and perform a final error correction on the error signal after the preliminary correction through the second error optimization factor to obtain a final error signal.

7. A method for controlling the ambient temperature of a substation communication machine room according to claim 6, characterized in that, According to the above, by evaluating the change of the temperature monitoring data of the substation communication room, the actual temperature change rate and the degree to which the actual temperature is close to the upper limit are obtained; by performing a correlation analysis on the recent equipment activity status of the actual temperature change rate, a correlation mark is obtained; By evaluating the normal fluctuation deviation of the actual temperature change rate, the actual temperature fluctuation deviation is obtained, including: Get the set temperature sampling period, and divide the difference between two consecutive temperature sampling monitoring values by the temperature sampling period to obtain the actual temperature change rate; The warning temperature threshold for triggering the correction mechanism and the maximum safe temperature allowed in the substation communication room are obtained. The calculation formula for the degree to which the actual temperature is close to the upper limit is: ； Among them, indicates the degree to which the actual temperature of the substation communication machine room is close to the upper limit at moment; indicates the actual monitored temperature of the substation communication machine room at moment; indicates the warning temperature threshold for triggering the correction mechanism; indicates the highest safe temperature allowed for the substation communication machine room. The time window set for the evaluation of the associated mark, the rapid temperature rise threshold, the high activity intensity threshold, the rapid change rate threshold, the acceleration threshold and the logic operation dual judgment condition are obtained. The evaluation process of the associated mark is to perform logical judgment through the logic operation function of the logic operation dual judgment condition. The first condition of the logic operation dual judgment condition is that the current actual temperature change rate exceeds the set rapid temperature rise threshold; the second condition of the logic operation dual judgment condition is whether there is a situation of high device activity or drastic dynamic change in the time window set for the evaluation of the associated mark. The judgment condition for the situation of high device activity or drastic dynamic change includes whether any of the following conditions are met at any time in the time window of the associated mark evaluation: Condition 1: The device aggregate activity level indicator is greater than the high activity intensity threshold; Condition 2: The rate of change of the device aggregate activity level indicator is greater than the rapid change rate threshold; Condition 3: The acceleration of the change of the device aggregate activity level indicator is greater than the acceleration threshold; When and only when both the first condition and the second condition are satisfied, the associated flag takes the value of , otherwise it takes the value of ; An acceptable upper limit of the temperature change rate is obtained, and a calculation result of subtracting the actual temperature change rate from the acceptable upper limit of the temperature change rate is used as the fluctuation deviation of the actual temperature.

8. A method for controlling the ambient temperature of a substation communication machine room according to claim 6, characterized in that, Based on the fusion evaluation of the degree of proximity of the actual temperature to the upper limit, the correlation flag, and the fluctuation deviation, a second error optimization factor is obtained, and the finally corrected error signal is obtained by finally correcting the preliminarily corrected error signal with the second error optimization factor, specifically including: Obtain the maximum value of the set second error optimization factor, and the calculation formula of the second error optimization factor is: ； Among them, represents the second error optimization factor of the substation communication machine room at moment; represents the set maximum value of the second error optimization factor; represents the correlation flag of the substation communication machine room at moment; represents the degree of proximity to the upper limit of the actual temperature of the substation communication machine room at moment; represents the fluctuation deviation of the actual temperature of the substation communication machine room at moment. Obtain the preliminarily corrected error signal, and use the calculation result of adding the preliminarily corrected error signal and the second error optimization factor as the finally corrected error signal.

9. The environmental temperature control method for a substation communication machine room according to claim 1, wherein, Based on the finally corrected error result, obtain the optimized temperature control output, specifically including: Use the finally corrected error signal as the input, apply it to a standard PID controller algorithm, perform proportional, integral, and differential operations on the input finally corrected error signal, obtain the adjustment control signal output by the PID controller algorithm, and use this adjustment control signal as the optimized temperature control output.

Images