Central air conditioner operation control method based on load prediction credibility determination

By using multi-timescale load forecasting and a two-level gating mechanism, and combining fluctuation, trend and margin information for arbitration decision-making, the problem of failing to predict future loads in traditional central air conditioning control has been solved, resulting in improved energy efficiency and extended equipment life.

CN122015261BActive Publication Date: 2026-06-19GUANGZHOU MINGHAN TECH CO LTD

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
GUANGZHOU MINGHAN TECH CO LTD
Filing Date
2026-04-14
Publication Date
2026-06-19

AI Technical Summary

Technical Problem

Traditional central air conditioning operation and control methods are based on real-time load feedback, which cannot effectively predict future loads, resulting in energy waste, impact on comfort, and equipment wear and tear. Furthermore, they do not fully consider the actual constraints of the system, leading to control oscillations and frequent equipment operation.

Method used

By employing multi-timescale load forecasting, forecast reliability assessment, and a two-level gating mechanism, and combining fluctuation, trend, and margin information for arbitration decisions, host start-up and shutdown constraints are set to achieve safe and flexible control.

Benefits of technology

It improves the energy efficiency of central air conditioning operation, reduces equipment wear and tear, enhances the safety and stability of control decisions, and extends the service life of equipment.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122015261B_ABST
    Figure CN122015261B_ABST
Patent Text Reader

Abstract

This invention provides a central air conditioning operation control method based on load forecast reliability determination, including data acquisition, multi-timescale load forecasting, forecast reliability assessment, load fluctuation characteristic analysis, forecast reliability gating, operation mode arbitration decision, hierarchical control execution and safety verification. Through reliable forecast-driven control, energy efficiency is improved and equipment wear is reduced while ensuring safety and comfort.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of central air conditioning operation control technology, and in particular to a central air conditioning operation control method based on load prediction reliability determination. Background Technology

[0002] Central air conditioning is a core energy-consuming device in large commercial buildings, office buildings, industrial parks, and other similar settings. The rationality of its operation and control directly affects system energy efficiency, equipment lifespan, and indoor comfort. Traditional central air conditioning operation and control methods are based on the current load, relying solely on real-time load feedback for adjustment. They trigger preset adjustments only by comparing real-time collected load-related parameters (such as temperature, pressure, and flow rate) with set thresholds, failing to predict future loads. This results in energy waste and delayed response to sudden load changes, impacting comfort.

[0003] Existing technologies incorporate load forecasting to achieve forward control of central air conditioning systems. However, existing predictive control schemes have significant technical drawbacks: First, they often use point forecasting to output single load values, failing to characterize the uncertainty of the forecast results and lacking a quality assessment mechanism. Blindly using erroneous forecasts can easily lead to system fluctuations, or even insufficient or excessive cooling. Second, control decisions rely solely on load fluctuation information without integrating key dimensions such as load trends and system capacity margins, which can result in aggressive or lagging control, leading to increased energy consumption or frequent equipment operation. Third, they fail to fully consider the actual engineering constraints of central air conditioning systems, such as cooling capacity transfer delays, system thermal inertia, and high losses during start-up and shutdown of the main unit. This disconnect between theoretical design and engineering implementation can cause control oscillations during command execution, significantly reducing equipment lifespan. Summary of the Invention

[0004] In view of this, the purpose of this invention is to provide a central air conditioning operation control method based on load prediction reliability determination, which improves energy efficiency and reduces equipment wear while ensuring safety and comfort through reliable prediction-driven control.

[0005] To solve the above-mentioned technical problems, the technical solution used in this invention is as follows:

[0006] The central air conditioning operation control method based on load forecast reliability determination according to the present invention includes the following steps:

[0007] S1. Data Acquisition: Acquire operational and environmental data related to the operation of the central air conditioning system.

[0008] S2. Multi-timescale load forecasting: Construct a multi-timescale load forecasting model to predict load changes at different time scales in the future based on operational and environmental data, and generate load forecasting intervals at different time scales.

[0009] S3. Forecast reliability assessment; calculate the overall forecast reliability of the load forecast interval.

[0010] S4. Load fluctuation characteristic analysis: Calculate the current fluctuation intensity of the load based on operation and environmental data, and extract future fluctuation tendency information based on the load forecast interval.

[0011] S5. Prediction reliability gating; including security gating and information gating; security gating, determines whether the overall prediction reliability is lower than the preset security threshold. If so, the central air conditioning is driven by the preset safe operation mode; if not, information gating is performed to determine whether future fluctuation tendency information is allowed to participate in control decisions; then proceed to S6.

[0012] S6. Operation mode arbitration decision: Introduce fluctuation information, trend information, and margin information as arbitration inputs, and select degradation arbitration or prospective arbitration based on the information gating results; Divide different working conditions and match the corresponding operation modes, and output a set of engineering actions including target set values, action constraints, and host start-stop prediction.

[0013] S7. Layered control execution and safety verification; establish equivalent time delay parameters for system cold capacity delivery, and perform layered execution according to time-scaled engineering action sets; set constraints for host start-up and shutdown, perform safety verification on control results, and correct instructions and return to safe operation mode when violations occur.

[0014] Preferably, the operation and environment data includes load data, environmental data, and time data; the load forecast intervals at different time scales include short-term load forecast intervals and medium-term load forecast intervals.

[0015] Preferably, S3 specifically includes the following steps:

[0016] S3.1 Calculate the reliability of each load forecast interval.

[0017] S3.2 Determine the consistency of different load forecast intervals.

[0018] S3.3. Combine the reliability of each load forecast interval and the consistency results of different load forecast intervals to obtain the comprehensive forecast reliability.

[0019] Preferably, S3.1 includes assessing whether each load forecast interval covers historical actual load values; assessing whether each load forecast interval has effective information; assessing whether each load forecast interval changes smoothly over time; and assessing whether there are abnormal jumps in the load forecast interval.

[0020] Preferably, S3.2 includes determining whether the short-term load forecast range falls within the medium-term load forecast range; and determining the degree of consistency between the short-term load forecast range and the medium-term load forecast range in terms of load change risk.

[0021] Preferably, the volatility information includes the current volatility intensity and future volatility tendency; the trend information includes short-term trend and medium-term trend; and the margin information is the capacity margin M.

[0022] Preferably, the degradation arbitration is a decision-making mode based solely on current real-time information, dividing the operating conditions into three categories: low-fluctuation steady state L1, medium-fluctuation transitional state L2, and high-fluctuation insufficient information L3. It only allows fine-tuning of the chilled pump frequency and chilled water outlet temperature, and prohibits the start-up and shutdown of the main unit.

[0023] The forward-looking arbitration integrates information on the current fluctuation intensity and future fluctuation tendency of the current load; and classifies the operating conditions into five categories: steady-state predictability (F1), storm warning (F2), continuous high fluctuation (F3), short-term disturbance impact (F4), and uncertain transition state (F5).

[0024] Preferably, the constraints for host start-up and shutdown include: minimum continuous running time constraint, minimum downtime constraint, multi-timescale prediction consistency constraint, and execution limit verification constraint. The host start-up and shutdown determination is only allowed when all constraints are satisfied. The content of the safety verification includes: whether the control command violates the equipment operating boundary, change rate constraint, and start-up and shutdown time constraint. If it violates the constraint, the corresponding command is automatically corrected or rejected, and the system reverts to the preset safe operation control mode.

[0025] Compared with existing technologies, the main advantages of the central air conditioning operation control method based on load prediction reliability determination described in this invention are as follows:

[0026] This method uses prediction reliability as the core entry threshold for look-ahead control, which solves the problem of blindly using prediction results in traditional predictive control. It avoids system operation fluctuations caused by prediction errors from the root and improves the safety of control decisions.

[0027] A two-tiered gating system is established using safety gating and information gating. In safety gating, a comparison is made between the safety threshold and the overall prediction reliability. If the overall prediction reliability is lower than the preset safety threshold, the prediction structure is deemed unusable, and a preset safe operating mode is used to avoid the impact of prediction errors on the central air conditioning operation. In information gating, it is further determined whether to use future fluctuation trend information in control decisions by assessing the consistency of different load prediction intervals. If the consistency of load change risk assessment across different load prediction intervals is lower than a preset value, future fluctuation trend information is not used in control decisions to reduce risk.

[0028] By integrating volatility information, trend information, and margin information for arbitration decision-making, and setting the margin information as the highest decision priority, the system avoids aggressive or lagging control caused by decision-making based on single information. At the same time, it selects between a degenerate arbitration mode or a forward-looking arbitration mode based on the credibility gating results, thus achieving a balance between the flexibility and robustness of decision-making.

[0029] This method fully considers the actual engineering constraints of central air conditioning systems, introduces time delay modeling to suppress control oscillations, divides the execution time scale into three levels, sets multiple strong constraints for host start-up and shutdown, avoids the disconnect between predictive control theory design and engineering implementation, significantly reduces the losses caused by frequent equipment operation, and extends the service life of core equipment. Attached Figure Description

[0030] The above and other objects, features, and advantages of the invention will become clearer through a more detailed description of the preferred embodiments illustrated in the accompanying drawings. The same reference numerals denote the same parts throughout the drawings, and the drawings are not intentionally drawn to scale with actual dimensions; the focus is on illustrating the gist of the invention.

[0031] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0032] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments, so that those skilled in the art can better understand the present invention and implement it. However, the embodiments are not intended to limit the present invention. In this embodiment, it should be understood that the terms "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the present invention and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting the present invention.

[0033] It should be noted that when one element is considered to be "connected" to another element, it can be directly connected to and integrated with the other element, or there may be an intervening element present. The terms "mounted," "one end," "the other end," and similar expressions used in this invention are for illustrative purposes only.

[0034] like Figure 1 As shown, a central air conditioning operation control method based on load forecast reliability determination includes the following steps:

[0035] S1. Data Acquisition: Acquire operational and environmental data for the central air conditioning system. This operational and environmental data includes load data, environmental data, and time data. The load data includes historical load time series data; the environmental data includes ambient temperature data and ambient humidity data; and the time data includes time-related data such as the current date, time period, and day of the week. In one embodiment, the data is cleaned and denoised.

[0036] S2. Multi-timescale load forecasting: Construct a multi-timescale load forecasting model to predict load changes at different time scales in the future based on operational and environmental data, and generate load forecasting intervals at different time scales, including short-term load forecasting intervals and medium-term load forecasting intervals.

[0037] S3. Forecast Reliability Assessment; Calculate the overall forecast reliability for the load forecast interval. This includes the following steps:

[0038] S3.1 Single-scale reliability calculation; calculate the reliability of each load forecast interval. This includes assessing whether each load forecast interval covers historical actual load; assessing whether each load forecast interval has effective information; assessing whether each load forecast interval has stability over time; and assessing whether there are abnormal jumps in the load forecast interval.

[0039] S3.2 Multi-scale consistency assessment; assessing the consistency of different load forecast intervals. This includes determining whether the short-term load forecast interval falls within the medium-term load forecast interval; and assessing the degree of consistency between the short-term and medium-term load forecast intervals in terms of load change risk.

[0040] S3.3. Combine the reliability of each load forecast interval and the consistency results of different load forecast intervals to obtain the comprehensive forecast reliability.

[0041] Specifically, coverage assessment: Based on the actual load data within a preset historical time window, calculate the coverage rate of the load forecast interval to the historical actual load; the historical time window is the continuous 7 to 30 days of operating data before the current moment; the ratio of the number of times the actual load value falls into the load forecast interval within the historical time window to the total number of statistical counts is used as the coverage rate; when the coverage rate is greater than the preset coverage rate threshold, the load forecast interval is determined to have coverage effectiveness.

[0042] Effective information content assessment: Calculate the relative ratio between the width of the load forecast interval and the center value of the forecast interval; where the interval width is the difference between the upper and lower boundaries of the forecast interval, and the interval center value is the average of the upper and lower boundaries of the forecast interval; when the ratio of the load forecast interval width to the interval center value is less than a preset width threshold, the load forecast interval is determined to have effective information content.

[0043] Stationarity assessment: Based on the changes in the upper and lower boundaries of the prediction interval at consecutive time points, the rate of change of the interval boundary at adjacent time points is calculated; the rate of change is the ratio of the change in the upper boundary of the interval at adjacent time points to the upper boundary of the interval at the previous time point, and the ratio of the change in the lower boundary of the interval at adjacent time points to the lower boundary of the interval at the previous time point; when the rate of change is less than a preset rate of change threshold, the change process of the prediction interval is determined to be stationary.

[0044] Abnormal jump assessment: When the rate of change of the boundary between adjacent time intervals is greater than the preset abnormal jump threshold, an abnormal jump is determined to exist.

[0045] S4. Load fluctuation characteristic analysis: Calculate the current fluctuation intensity of the load based on operation and environmental data, and extract future fluctuation tendency information based on the load forecast interval.

[0046] S5. Prediction reliability gating; including security gating and information gating; security gating, determines whether the overall prediction reliability is lower than the preset security threshold. If so, the central air conditioning is driven by the preset safe operation mode; if not, information gating is performed to determine whether future fluctuation tendency information is allowed to participate in control decisions; then proceed to S6.

[0047] The determination of whether to allow future fluctuation tendency information to participate in control decision-making specifically involves: pre-setting a determination of whether to allow future fluctuation tendency information to participate in control decision-making, and determining whether the comprehensive prediction credibility is greater than or equal to the information gating threshold. If yes, then future fluctuation tendency information is allowed to participate in control decision-making; if no, then future fluctuation tendency information is not allowed to participate in control decision-making.

[0048] S6. Operation mode arbitration decision: Introduce fluctuation information, trend information, and margin information as arbitration inputs, and select degradation arbitration or prospective arbitration based on the information gating results; Divide different working conditions and match the corresponding operation modes, and output a set of engineering actions including target set values, action constraints, and host start-stop prediction.

[0049] Specifically, when the overall forecast credibility is greater than or equal to the information gating threshold, future volatility tendency information is allowed to participate in control decisions, thus choosing forward arbitration; when the overall forecast credibility is less than the information gating threshold, future volatility tendency information is allowed to participate in control decisions, thus choosing degradation arbitration.

[0050] S7. Layered control execution and safety verification; establish equivalent time delay parameters for system cold capacity delivery, and perform layered execution according to time-scaled engineering action sets; set constraints for host start-up and shutdown, perform safety verification on control results, and correct instructions and return to safe operation mode when violations occur.

[0051] The volatility information includes the current volatility intensity and future volatility tendency; the trend information includes short-term and medium-term trends; and the margin information is the capacity margin M.

[0052] Short-term trend Ds is used to determine the direction of pump frequency and outlet water temperature; medium-term trend Dm is used to determine the start-up and shutdown direction of the main unit; short-term and medium-term trends are characterized as upward, downward, or stable trends.

[0053] The degradation arbitration is a decision-making mode based solely on current real-time information, dividing the operating conditions into three categories: low-fluctuation steady state (L1), medium-fluctuation transitional state (L2), and high-fluctuation insufficient information (L3). It only allows fine-tuning of the chilled pump frequency and chilled water outlet temperature, and prohibits the start-up and shutdown of the main unit.

[0054] The low-fluctuation steady-state L1 condition refers to a state where the current fluctuation intensity is low and the load change is stable. Its characteristics are: within multiple consecutive control cycles, the real-time load change amplitude is small, the change rate is less than the preset low-fluctuation threshold, and there is no obvious sudden increase or decrease phenomenon. The system is operating in a stable range. Its mode is conservative, and the engineering actions are: keep the current number of host machines running unchanged; maintain the current system operating state as the main focus.

[0055] Based on the direction of the short-term trend Ds, the frequency of the chilled water pump is slightly adjusted; based on the direction of the short-term trend Ds, the chilled water outlet temperature is finely adjusted to maintain stable system operation and avoid unnecessary adjustments.

[0056] The L2 transitional operating condition with moderate fluctuations refers to a state where current fluctuations exhibit some variation but remain within a controllable range overall. Its characteristics include: over multiple consecutive control cycles, the real-time load variation amplitude is within a moderate range, with the rate of change falling between a preset low-fluctuation threshold and a moderate-fluctuation threshold, or exhibiting a slow upward or downward trend; the system is in a transitional adjustment phase. Its mode is "following," and the engineering actions are: maintaining the current host operating state unchanged; while maintaining overall system stability, appropriately adjusting the chilled water pump frequency according to the direction of change in the short-term trend Ds to gradually match the system's cooling capacity with the current load variation trend; and making small dynamic adjustments to the chilled water outlet temperature according to the direction of change in the short-term trend Ds to achieve follow-up adjustment of load changes, buffering system fluctuations and maintaining stable system operation.

[0057] The high-fluctuation information deficiency L3 condition refers to a state where the current fluctuation is large or the information is unstable. Its characteristics include: large real-time load changes over multiple consecutive control cycles, with a change rate exceeding the preset medium fluctuation threshold, or the presence of significant sudden increases or decreases, resulting in high system operational uncertainty. The mode is buffer / sensitive, and the engineering actions are: maintaining the current number of operating main units unchanged, prohibiting the addition or shutdown of main units; rapidly adjusting the chilled water pump frequency based on the current real-time load changes to improve the system's short-term response capability; and appropriately reducing the chilled water outlet temperature based on the current real-time load changes to increase the system's cooling margin, thereby buffering the impact of sudden load changes on system operational stability.

[0058] The forward-looking arbitration integrates information on the current fluctuation intensity and future fluctuation tendency of the current load; and classifies the operating conditions into five categories: steady-state predictability (F1), storm warning (F2), continuous high fluctuation (F3), short-term disturbance impact (F4), and uncertain transition state (F5).

[0059] The steady-state predictable F1 condition has low current fluctuation intensity and low future fluctuation tendency. Its mode is conservative, and the engineering actions are: maintain the current number of main units in operation unchanged; make a small adjustment to the chilled water pump frequency according to the direction of the short-term trend Ds; and make a fine adjustment to the chilled water outlet temperature according to the direction of the short-term trend Ds to maintain the stable operation of the system and avoid unnecessary frequent adjustments.

[0060] The storm warning F2 condition indicates low current volatility but a high potential for future volatility. Its mode is: prepared and sensitive; the operational actions are: while maintaining the current mainframe operating status, based on the direction of the medium-term trend Dm, to determine in advance whether to increase the number of operating mainframes; and based on the changing direction of the short-term trend Ds...

[0061] By appropriately increasing the frequency of the chilled water pump and appropriately decreasing the chilled water outlet temperature, the system's cooling capacity can be increased in advance.

[0062] Prepare for potential future load increases.

[0063] The continuous high-fluctuation F3 operating condition is characterized by high current fluctuation intensity and a high future fluctuation tendency. Its mode is: sensitive, and the engineering actions are: adjust the number of main units in operation in a timely manner according to the change direction of the medium-term trend Dm to match the continuously high load demand; and dynamically adjust the chilled water pump frequency and appropriately reduce the chilled water outlet temperature according to the change direction of the short-term trend Ds to ensure that the system has sufficient cooling capacity and maintains stable system operation.

[0064] The short-term disturbance impact F4 condition is characterized by high current fluctuation intensity and low future fluctuation tendency. Its mode is buffering, and the engineering actions are as follows: maintain the current host operating state unchanged to avoid unnecessary host start-up and shutdown caused by short-term disturbances; quickly adjust the chilled water pump frequency and appropriately reduce the chilled water outlet temperature according to the changing direction of the short-term trend Ds to enhance the system's short-term buffering capacity and absorb instantaneous load impacts.

[0065] The uncertain transition state F5 is a state where the current fluctuation trend is unclear or the information is insufficient. Its mode is: observation, and the engineering actions are: keep the current host operating status unchanged; mainly maintain the current chilled pump frequency and chilled water outlet temperature; make small dynamic adjustments according to the real-time load changes, in order to observe the system change trend and avoid over-adjustment or malfunctions under uncertain information conditions.

[0066] The timescale includes fast, medium, and slow timescales. The fast timescale is used for chilled water pump frequency regulation; its control cycle is short, allowing for rapid adjustment of hydraulic delivery capacity. The medium timescale is used for chilled water outlet temperature setting and regulation; its control cycle is relatively long to adapt to the thermal inertia characteristics of the chiller and system. The slow timescale is used for chiller start-up and shutdown control; its control cycle is significantly longer than that of the fast and medium timescales.

[0067] The constraints for host start-up and shutdown include: minimum continuous running time constraint, minimum downtime constraint, multi-timescale prediction consistency constraint, and execution limit verification constraint. Host start-up and shutdown determination is only allowed when all constraints are satisfied. The content of the safety verification includes: whether the control command violates the equipment operating boundary, change rate constraint, and start-up and shutdown time constraint. If it violates the constraint, the corresponding command is automatically corrected or rejected, and the system reverts to the preset safe operation control mode.

[0068] The following examples illustrate this:

[0069] The central air conditioning system in this embodiment serves a high-rise office building. The system is equipped with three chiller units and variable frequency chilled water pumps. The operating environment is characterized by significant load fluctuations on weekdays, with sharp increases in load during the morning and evening peak hours, a decrease in load during the lunch break, and a significant reduction in load on weekends. It is also susceptible to sudden load changes due to outdoor weather conditions. The specific steps are as follows:

[0070] S1. Collect operational and environmental data of the central air conditioning system over the past 6 months.

[0071] S2. Set the short-term forecast time scale to 30 minutes and the medium-term forecast time scale to 3 hours. Use the multi-time-scale load forecasting model to forecast short-term and medium-term load changes; generate short-term load forecast intervals and medium-term load forecast intervals.

[0072] S3.1 Single-scale reliability calculation: The reliability of the short-term load forecast interval and the medium-term load forecast interval are calculated respectively. The short-term load forecast interval has a coverage rate of 92% over the historical actual load, an interval width of 5%-8%, and no abnormal jumps. The medium-term load forecast interval has a coverage rate of 90% over the historical actual load, an interval width of 8%-12%, and no abnormal jumps. The reliability of both the short-term load forecast interval and the medium-term load forecast interval is determined to be excellent.

[0073] S3.2 Multi-scale consistency judgment: The short-term load forecast intervals all reasonably fall within the corresponding time periods of the medium-term load forecast intervals; the loads of both show an upward trend, the load change risk judgments are consistent, and the multi-scale consistency is judged to be excellent.

[0074] S3.3. The single-scale confidence result and the multi-scale consistency judgment are combined and integrated to calculate the comprehensive prediction confidence of 0.91.

[0075] S4. Based on historical load time series data, the current load fluctuation intensity (Vcur) is calculated to be 0.15, which is lower than the low fluctuation threshold of 0.2, and the current state is determined to be low fluctuation. Based on the short-term load forecast interval, the future load fluctuation tendency (Vcur) for the next 30 minutes is extracted to be 0.85, which is higher than the high fluctuation threshold of 0.7, and the risk of high fluctuation in the future is determined to exist.

[0076] S5. The preset safety threshold is 0.7; the preset information gating threshold is 0.8; the comprehensive prediction confidence is 0.91, which is greater than the safety threshold; then it is determined whether the comprehensive prediction confidence is greater than or equal to the information gating threshold. If the comprehensive prediction confidence is 0.91, which is greater than the information gating threshold, future fluctuation tendency information is allowed to participate in control decisions.

[0077] S6. Select the forward-looking arbitration mode; the current fluctuation intensity is low, and the future fluctuation tendency is high. Classify the operating condition as storm pre-warning condition F2, and use the pre-sensitive operation mode. Working action output: chilled water outlet temperature pre-drops by 0.5℃ / step, chilled water pump frequency pre-increases by 1Hz / step, triggers start-up prediction calculation, and temporarily does not execute the start-up action. The action constraints are a minimum outlet water temperature of 7℃ and a maximum pump frequency of 50Hz.

[0078] S7. Time Delay Modeling: Based on the system pipeline length, set the equivalent time delay τ_delay for cold energy delivery to 5 minutes. After the control action is executed, suppress the reverse correction action within 5 minutes.

[0079] Layered execution: The fast timescale executes a 1Hz pre-increase in chilled pump frequency, the medium timescale executes a 0.5℃ pre-decrease in chilled water outlet temperature, the upper-layer module outputs action constraints, and the lower-layer fuzzy PI controller achieves smooth adjustment.

[0080] Main unit start / stop constraint verification: Although the current trend is upward and the margin is close to the boundary, the execution limit has not been reached (the outlet water temperature has not reached the lower limit and the pump frequency has not reached the upper limit), the start-up constraint is not met, and only the start-up prediction is maintained.

[0081] Safety verification: The executed instructions did not violate the device's operating boundaries and rate of change constraints, and the system is operating normally.

[0082] In another embodiment, which is a weekend with overcast and rainy weather, the overall load of the office building's central air conditioning system is low, and the sudden drop in outdoor temperature causes a significant deviation in the load forecast results; specifically, the following steps are included:

[0083] S1. Collect operational and environmental data of the central air conditioning system over the past 6 months.

[0084] S2. Set the short-term forecast time scale to 30 minutes and the medium-term forecast time scale to 3 hours. Use the multi-time-scale load forecasting model to forecast short-term and medium-term load changes; generate short-term load forecast intervals and medium-term load forecast intervals.

[0085] S3.1 Single-scale reliability calculation; calculate the reliability of the short-term load forecast interval and the medium-term load forecast interval respectively; the coverage rate of the short-term load forecast interval to the historical actual load is 65%; the coverage rate of the medium-term load forecast interval to the historical actual load is 70%, and the reliability of both the short-term load forecast interval and the medium-term load forecast interval is determined to be poor.

[0086] S3.2 Multi-scale consistency judgment: The short-term load forecast range increases slightly, while the medium-term load forecast range decreases. The load trends of the two are different, and the judgment of load change risk is different. Therefore, the multi-scale consistency is judged as poor.

[0087] S3.3. The single-scale confidence result and the multi-scale consistency judgment are combined and integrated to calculate the comprehensive prediction confidence of 0.55.

[0088] S4. Based on historical load time series data, the current load fluctuation intensity (Vcur) is calculated to be 0.15, which is lower than the low fluctuation threshold of 0.2. Therefore, the current state is determined to be low fluctuation.

[0089] S5. The overall prediction confidence level is 0.55, which is lower than the preset safety threshold of 0.7; therefore, the preset safe operation mode is used to drive the operation of the central air conditioning.

[0090] S6. The open-circuit system is divided into low-fluctuation steady-state L1 operating conditions. A conservative operating mode is adopted, with the chilled water outlet temperature increased by 0.5℃ and the chilled water pump frequency decreased by 1Hz to maximize energy efficiency.

[0091] S7. Strictly adhere to action constraints and prohibit all host start-up and shutdown operations.

[0092] In this embodiment, when the reliability of the comprehensive prediction is unreliable, the system is kept running stably through a safe operation mode, while energy saving is achieved under low load conditions.

[0093] This method uses prediction reliability as the core entry threshold for look-ahead control, which solves the problem of blindly using prediction results in traditional predictive control. It avoids system operation fluctuations caused by prediction errors from the root and improves the safety of control decisions.

[0094] A two-tiered gating system is established using safety gating and information gating. In safety gating, a comparison is made between the safety threshold and the overall prediction reliability. If the overall prediction reliability is lower than the preset safety threshold, the prediction structure is deemed unusable, and a preset safe operating mode is used to avoid the impact of prediction errors on the central air conditioning operation. In information gating, it is further determined whether to use future fluctuation trend information in control decisions by assessing the consistency of different load prediction intervals. If the consistency of load change risk assessment across different load prediction intervals is lower than a preset value, future fluctuation trend information is not used in control decisions to reduce risk.

[0095] By integrating volatility information, trend information, and margin information for arbitration decision-making, and setting the margin information as the highest decision priority, the system avoids aggressive or lagging control caused by decision-making based on single information. At the same time, it selects between a degenerate arbitration mode or a forward-looking arbitration mode based on the credibility gating results, thus achieving a balance between the flexibility and robustness of decision-making.

[0096] This method fully considers the actual engineering constraints of central air conditioning systems, introduces time delay modeling to suppress control oscillations, divides the execution time scale into three levels, sets multiple strong constraints for host start-up and shutdown, avoids the disconnect between predictive control theory design and engineering implementation, significantly reduces the losses caused by frequent equipment operation, and extends the service life of core equipment.

[0097] In this specification, unless otherwise expressly specified and limited, "above" or "below" the second feature can mean that the first and second features are in direct contact, or that the first and second features are in indirect contact through an intermediate medium. Furthermore, "above," "over," and "on top" of the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply that the first feature is at a lower horizontal level than the second feature.

[0098] In the description of this specification, the references to terms such as "preferred embodiment," "another embodiment," "other embodiment," or "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in a suitable manner in any one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0099] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A central air conditioning operation control method based on load forecast reliability determination, characterized in that: Includes the following steps: S1. Data Acquisition; Acquiring operational and environmental data related to the central air conditioning system; S2. Multi-timescale load forecasting: Construct a multi-timescale load forecasting model to predict load changes at different time scales in the future based on operational and environmental data, and generate load forecasting intervals at different time scales. S3. Forecast reliability assessment; calculate the overall forecast reliability for the load forecast interval; S3.1 Calculate the reliability of each load forecast interval; including assessing whether each load forecast interval covers historical actual load values; assessing whether each load forecast interval has effective information; assessing whether each load forecast interval has stability over time; and assessing whether there are abnormal jumps in the load forecast interval. S3.2 Determine the consistency between different load forecast intervals; including determining whether the short-term load forecast interval falls within the medium-term load forecast interval; and determining the degree of consistency between the short-term and medium-term load forecast intervals in terms of load change risk. S3.

3. Combine the reliability of each load forecast interval and the consistency results of different load forecast intervals to obtain the comprehensive forecast reliability. S4. Load fluctuation characteristic analysis: Calculate the current fluctuation intensity of the load based on operation and environmental data, and extract future fluctuation tendency information based on the load forecast interval; S5. Prediction reliability gating; including security gating and information gating; security gating determines whether the overall prediction reliability is lower than the preset security threshold. If so, the central air conditioning is driven by the preset safe operation mode; if not, information gating is performed to determine whether to allow future fluctuation tendency information to participate in control decisions. Then proceed to S6; S6. Arbitration decision-making in operation mode; introduce fluctuation information, trend information, and margin information as arbitration inputs, and select degenerate arbitration or prospective arbitration based on the information gating results; Divide different working conditions and match the corresponding operating modes, and output a set of engineering actions including target set values, action constraints, and host start / stop predictions; S7. Layered control execution and security verification; Establish equivalent time delay parameters for system cold capacity delivery, and perform layered execution according to time-scaled engineering action sets; set constraints for host start-up and shutdown, perform safety checks on control results, and correct instructions and return to safe operation mode when violations occur.

2. The central air conditioning operation control method based on load forecast reliability determination according to claim 1, characterized in that: The operational and environmental data include load data, environmental data, and time data; the load forecast intervals at different time scales include short-term load forecast intervals and medium-term load forecast intervals.

3. The central air conditioning operation control method based on load forecast reliability determination according to claim 2, characterized in that: The fluctuation information includes the current fluctuation intensity and the future fluctuation trend; The trend information includes short-term trends and medium-term trends; The margin information is the capability margin M.

4. The central air conditioning operation control method based on load forecast reliability determination according to claim 1, characterized in that: The degradation arbitration is a decision-making mode based solely on current real-time information, which divides the operating conditions into three categories: low-fluctuation steady state L1, medium-fluctuation transitional state L2, and high-fluctuation insufficient information L3. It only allows fine-tuning of the chilled pump frequency and chilled water outlet temperature, and prohibits the start-up and shutdown of the main unit. The forward-looking arbitration integrates information on the current fluctuation intensity and future fluctuation tendency of the current load; and classifies the operating conditions into five categories: steady-state predictability (F1), storm warning (F2), continuous high fluctuation (F3), short-term disturbance impact (F4), and uncertain transition state (F5).

5. The central air conditioning operation control method based on load forecast reliability determination according to claim 1, characterized in that: The constraints for host start-up and shutdown include: minimum continuous running time constraint, minimum downtime constraint, multi-timescale prediction consistency constraint, and execution limit verification constraint. Host start-up and shutdown determination is only allowed when all constraints are satisfied. The content of the safety verification includes: whether the control command violates the equipment operating boundary, change rate constraint, and start-up and shutdown time constraint. If it violates the constraint, the corresponding command is automatically corrected or rejected, and the system reverts to the preset safe operation control mode.