Method and system for predicting energy consumption of a pharmaceutical warehouse

Abstract

The application provides a kind of medical warehouse energy consumption prediction analysis method and system, belong to energy consumption management technical field, specifically include: with the opening data of attention equipment combination in different time periods, determine the matching environment temperature interval corresponding to attention equipment combination, according to the coincidence degree of the matching environment temperature interval corresponding to different attention equipment combinations and the number of attention equipment combinations, determine the adjustment processing strategy of attention equipment combination in different outdoor environment temperature interval, determine the attention equipment combination that is forced to adjust and process in different outdoor environment temperature interval based on adjustment processing strategy, and it is used as influence equipment combination, with the degree of association of influence equipment combination in different outdoor environment temperature interval and temperature adjustment processing data, the determination of energy consumption analysis method of energy consumption prediction model, lay the foundation for reducing energy consumption in temperature adjustment process.

Description

Technical Field

[0001] This invention belongs to the field of energy management technology, and in particular relates to a method and system for predicting and analyzing energy consumption in pharmaceutical warehouses. Background Technology

[0002] The energy consumption of pharmaceutical warehouses (including cool storage, cold storage, and frozen storage) mainly consists of refrigeration systems, dehumidification systems, lighting systems, and ventilation systems, with refrigeration and dehumidification typically accounting for over 70% of energy consumption. Unlike ordinary warehouses, the energy consumption of pharmaceutical warehouses is affected by a combination of factors: external environment (temperature and humidity, season, diurnal temperature range), internal load (drug storage volume, personnel workload, forklift operation), equipment status (refrigeration unit efficiency, dehumidifier performance, insulation effect), and the strict limitations imposed by drug storage regulations on temperature and humidity fluctuations.

[0003] Specifically, in order to predict energy consumption, the invention patent application CN202211451027.3, "A Method and Device for Energy Consumption Monitoring and Prediction," obtains ideal energy consumption data for the warehouse from the model output. This data is used to monitor the warehouse's energy management based on both ideal and actual energy consumption data. Furthermore, it allows for more accurate monitoring of energy management based on actual energy consumption data, order volume, regional attributes, and environmental data. However, this approach has the following drawbacks: In the process of temperature regulation in pharmaceutical warehouses, existing technical solutions often rely on fixed regulation methods. However, external environmental interference leads to differences in energy consumption among different combinations of temperature regulation equipment. Therefore, determining the energy consumption prediction model update analysis and processing strategy based on the degree of energy consumption anomalies during temperature regulation, as well as the regulation and control methods for different equipment combinations, to ensure reliable and comprehensive updates of the energy consumption prediction model and reduce energy consumption during temperature regulation, has become an urgent technical problem to be solved.

[0004] In summary, to solve the above-mentioned technical problems, this application provides a method and system for predicting and analyzing energy consumption in pharmaceutical warehouses. Summary of the Invention

[0005] To achieve the objectives of this invention, the following technical solution is adopted: Specifically, this application provides a method for predicting and analyzing energy consumption in pharmaceutical warehouses, which includes: S1 determines the distribution data of abnormal energy consumption periods in the pharmaceutical warehouse based on the energy consumption data of the pharmaceutical warehouse, and determines the update analysis and processing strategy of the energy consumption prediction model based on the distribution data of the abnormal energy consumption periods. S2 determines the combination of devices of interest in the temperature regulation device combination based on the update analysis and processing strategy and combined with the activation data of the temperature regulation device in different time periods. S3 uses the activation data of the monitored device combination in different time periods to determine the matching ambient temperature range corresponding to the monitored device combination. Based on the degree of overlap of the matching ambient temperature ranges corresponding to different monitored device combinations and the number of monitored device combinations, the adjustment and processing strategy of the monitored device combination in different outdoor ambient temperature ranges is determined. S4 determines the combination of devices of interest for forced regulation within different outdoor ambient temperature ranges based on the regulation and processing strategy, and uses these as the combination of devices of interest. Based on the combination of devices of interest within different outdoor ambient temperature ranges and the degree of correlation with temperature regulation and processing data, the energy consumption analysis method of the energy consumption prediction model is determined.

[0006] The beneficial effects of this invention are as follows: Based on the distribution data of abnormal energy consumption periods, the degree of energy consumption abnormality when using a combination of temperature control equipment for temperature control is determined. Based on the degree of energy consumption abnormality, the update analysis and processing strategy of the energy consumption prediction model is determined, thus laying the foundation for ensuring the reliable updating of the energy consumption prediction model and thus ensuring energy consumption control during the temperature control process.

[0007] Based on the activation data of temperature control equipment combinations at different times, the matching ambient temperature ranges corresponding to different temperature control equipment combinations are determined. Using the matching ambient temperature ranges corresponding to different temperature control equipment combinations and the update analysis and processing strategy, the degree of adaptation of the temperature control data update of different temperature control equipment combinations under the current update analysis and processing strategy is determined. Based on the degree of adaptation of the temperature control data update of different temperature control equipment combinations, the equipment combinations to be focused on are determined, that is, the temperature control equipment combinations that are forcibly activated within certain outdoor ambient temperature ranges are determined, thereby further improving the efficiency and comprehensiveness of the energy consumption prediction model update processing.

[0008] Furthermore, the energy consumption data of the pharmaceutical warehouse includes the energy consumption during the temperature regulation process at different times.

[0009] Furthermore, the abnormal energy consumption period of the pharmaceutical warehouse is determined based on the period when the energy consumption of the pharmaceutical warehouse exceeds a preset energy consumption threshold.

[0010] Furthermore, the method for determining the update analysis and processing strategy of the energy consumption prediction model is as follows: S11 determines the duration percentage of abnormal energy consumption periods on different dates based on the distribution data of the abnormal energy consumption periods. S12 determines the days with abnormal energy consumption based on the percentage of time periods with abnormal energy consumption in different dates; S13 uses the distribution data of the energy consumption anomaly dates to determine the update analysis and processing strategy for the energy consumption prediction model.

[0011] Furthermore, the method for determining the combination of devices of interest in the temperature control device combination is as follows: S21 uses the activation data of the temperature regulating devices in different time periods to determine the time periods during which the temperature regulating devices are activated, and uses the time periods during which the temperature regulating devices are activated as the matching regulation time periods of the temperature regulating devices. S22 determines the matching ambient temperature range corresponding to the temperature control device combination based on the matching adjustment period of the temperature control device combination; S23 determines the combination of devices of interest in the combination of temperature control devices based on the update analysis and processing strategy and the matching ambient temperature ranges corresponding to different combinations of temperature control devices.

[0012] Secondly, the present invention provides a computer system comprising: a memory and a processor connected in communication, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the above-described method for predicting and analyzing energy consumption in a pharmaceutical warehouse when running the computer program.

[0013] Other features and advantages will be set forth in the following description, and the objects and other advantages of the invention are realized and obtained through the structures particularly pointed out in the description and the drawings.

[0014] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0015] The above and other features and advantages of the present invention will become more apparent from a detailed description of exemplary embodiments thereof with reference to the accompanying drawings.

[0016] Figure 1 This is a flowchart of a method for predicting and analyzing energy consumption in a pharmaceutical warehouse. Figure 2 This is a flowchart illustrating the method for determining the update analysis and processing strategy of the energy consumption prediction model; Figure 3 This is a flowchart illustrating the method for determining the combination of devices of interest in a temperature control equipment assembly. Figure 4 This is a flowchart illustrating a method for determining the adjustment and processing strategies of equipment combinations within different outdoor ambient temperature ranges. Detailed Implementation

[0017] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments of this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0018] Example 1 like Figure 1 As shown, this application provides a method for predicting and analyzing energy consumption in pharmaceutical warehouses, specifically including: S1 determines the distribution data of abnormal energy consumption periods in the pharmaceutical warehouse based on the energy consumption data of the pharmaceutical warehouse, and determines the update analysis and processing strategy of the energy consumption prediction model based on the distribution data of the abnormal energy consumption periods. The energy consumption data of the pharmaceutical warehouse refers to the energy consumption records of the pharmaceutical warehouse during temperature regulation processes in different time periods, including a structured data sequence with a time dimension (the start and end times of a specific time period) and an energy consumption dimension (the total electrical energy consumed by the temperature regulation equipment during that time period). The abnormal energy consumption period refers to the period when the energy consumption of the pharmaceutical warehouse exceeds a preset energy consumption threshold, which is determined based on the energy consumption baseline level for normal temperature control operation of the pharmaceutical warehouse. The distribution data of abnormal energy consumption periods refers to the distribution information of the frequency and duration of abnormal energy consumption periods on different dates after classifying and statistically analyzing all abnormal energy consumption periods by date. The energy consumption prediction model refers to a predictive model used to predict the energy consumption required for temperature regulation in the pharmaceutical warehouse under different outdoor and indoor temperature conditions using different combinations of temperature regulation equipment. The update analysis and processing strategy refers to the management rules that determine under what data accumulation conditions the energy consumption prediction model will be updated based on the degree of energy consumption anomaly.

[0019] Assuming the pharmaceutical warehouse has accumulated several days of energy consumption monitoring data, with energy consumption records for multiple time periods each day; if energy consumption exceeds a preset energy consumption threshold during certain time periods on a certain day, these time periods are marked as abnormal energy consumption periods; by statistically analyzing the duration of abnormal energy consumption periods on each date, the number and distribution of abnormal energy consumption dates can be identified; this information then drives the hierarchical decision-making of the update analysis and processing strategy, determining the triggering conditions for model updates.

[0020] This step is the starting point of the entire method. Its significance lies in integrating discrete energy consumption data over time periods into energy consumption anomaly distribution characteristics based on dates. This enables subsequent model update strategy decisions to make hierarchical judgments based on the overall trend of energy consumption anomalies, avoiding excessive model updates due to local single-period anomalies, and also avoiding insufficient model update frequency due to underestimation of the degree of anomalies. This achieves a precise match between the timing of model updates and the degree of energy consumption anomalies.

[0021] Furthermore, the energy consumption data of the pharmaceutical warehouse includes the energy consumption during the temperature regulation process at different times.

[0022] Specifically, the abnormal energy consumption period of the pharmaceutical warehouse is determined based on the period when the energy consumption of the pharmaceutical warehouse exceeds a preset energy consumption threshold.

[0023] The preset energy consumption threshold refers to the upper limit reference value of energy consumption obtained from the statistical analysis of historical normal temperature control operation data of the pharmaceutical warehouse, which is used to distinguish between normal energy consumption periods and abnormal energy consumption periods.

[0024] Assuming that the energy consumption of a pharmaceutical warehouse under normal operating conditions is usually within a reasonable range every 15 minutes, the energy consumption threshold exceeding the upper limit of this range is set as the preset energy consumption threshold; if the total energy consumption of the temperature control equipment exceeds this threshold within a certain 15-minute period, then this period is marked as an abnormal energy consumption period.

[0025] This step provides an objective quantitative basis for identifying periods of abnormal energy consumption. Its significance lies in replacing manual experience-based judgment with a data-driven approach, ensuring the consistency and reproducibility of the results of abnormal energy consumption identification.

[0026] It should be noted that, as Figure 2 As shown, the method for determining the update analysis and processing strategy of the energy consumption prediction model is as follows: In this embodiment, based on the distribution data of abnormal energy consumption periods, the degree of energy consumption abnormality when using a combination of temperature control equipment for temperature control is determined. Based on the degree of energy consumption abnormality, the update analysis and processing strategy of the energy consumption prediction model is determined, thereby laying the foundation for ensuring the reliable updating of the energy consumption prediction model and thus ensuring energy consumption control during the temperature control process.

[0027] S11 determines the duration percentage of abnormal energy consumption periods on different dates based on the distribution data of the abnormal energy consumption periods. The percentage of abnormal energy consumption periods on different dates refers to the ratio of the total duration of abnormal energy consumption periods on a given date to the total duration of all temperature regulation periods on that date, with a value ranging from zero to one.

[0028] Suppose there are multiple temperature regulation periods on a certain day. During some of these periods, the energy consumption exceeds a preset energy consumption threshold and is marked as an abnormal energy consumption period. Then, the ratio of the total duration of the abnormal period to the total duration of all regulation periods on that day is calculated to obtain the percentage of the abnormal energy consumption period on that day.

[0029] This step aggregates the abnormal states of each time period into a date-level percentage indicator. Its significance lies in eliminating the impact of different adjustment periods on different dates in a standardized proportion, so that the degree of energy consumption anomalies between different dates has a unified dimension that can be directly compared.

[0030] S12 determines the days with abnormal energy consumption based on the percentage of time periods with abnormal energy consumption in different dates; The abnormal energy consumption date refers to a date where the duration of the abnormal energy consumption period is greater than a preset duration percentage threshold, indicating that most of the adjustment period on that date is in a high energy consumption state.

[0031] Assuming the preset time percentage threshold is set to a certain percentage, if the percentage of abnormal energy consumption periods on a certain day exceeds this threshold, then that day will be marked as an abnormal energy consumption day; otherwise, it will not be marked.

[0032] This step transforms the date-level percentage data into binary abnormal state labels. Its significance lies in establishing clear boundaries for judging abnormal dates, providing clear input conditions for the hierarchical decision-making of subsequent energy consumption prediction model update strategies.

[0033] Specifically, the dates with abnormal energy consumption are those where the duration of the abnormal energy consumption period is greater than a preset duration percentage threshold.

[0034] S13 uses the distribution data of the energy consumption anomaly dates to determine the update analysis and processing strategy for the energy consumption prediction model.

[0035] The distribution data of abnormal energy consumption dates refers to the marking of abnormal energy consumption status on each date within the statistical period, as well as the proportion of the number of abnormal energy consumption dates in all statistical dates.

[0036] Assuming the statistical period is the past 30 days, and the number of days marked as abnormal energy consumption is a certain number, the proportion of abnormal energy consumption days in the 30 days is calculated. Based on the relationship between this proportion and the preset threshold for the proportion of abnormal days, the corresponding update analysis and processing strategies are determined.

[0037] This step provides a basis for the final determination of the update analysis and processing strategy. Its significance lies in transforming the extent of energy consumption anomalies (the proportion of anomaly dates) into a quantitative representation of the model update frequency requirement. A higher proportion of anomalies corresponds to a more lenient update triggering condition, while a lower or zero anomaly corresponds to a more conservative update triggering condition, thereby achieving an adaptive match between update frequency and the degree of energy consumption risk.

[0038] It is understood that, based on the distribution data of the energy consumption anomaly dates, the update analysis and processing strategy for the energy consumption prediction model is determined, specifically including: Case 1: If there are no abnormal energy consumption dates, the energy consumption is low at this time, and the need for energy consumption regulation using the energy consumption prediction model is low. In this case, the update strategy of the energy consumption prediction model is determined to be a conservative update strategy. That is, when the number of updates for the regulation period in the temperature regulation equipment combination of at least the target combination number is greater than the preset update period number threshold, the energy consumption prediction model is updated. The conservative update strategy refers to requiring that, under the condition of lowest energy consumption anomaly, model updates are only triggered after a certain amount of data from multiple temperature control device combinations during their regulation periods have accumulated, thus avoiding frequent updates under low energy consumption risk. The target combination number refers to the minimum number of temperature control device combinations that must simultaneously meet the update conditions to trigger the conservative update strategy.

[0039] Assuming there are no abnormal energy consumption dates in the past 30 days, it indicates that the current energy consumption control is good and the energy consumption prediction model does not need to be updated frequently. In this case, a model update will only be triggered when the number of updates for the adjustment periods of each of the target combination of temperature control equipment combinations exceeds the preset threshold for the number of update periods.

[0040] This step represents the lowest demand scenario in the update analysis and processing strategy. Its significance lies in reducing unnecessary model update frequency by raising the update trigger threshold when energy consumption is stable, thereby saving computing resources and avoiding model parameter oscillations caused by excessive updates.

[0041] Case 2: If there are abnormal energy consumption dates, and the proportion of abnormal energy consumption dates is above the preset abnormal date proportion threshold, then the update processing strategy of the energy consumption prediction model is determined to be a relaxed update strategy. When the number of updates for the adjustment period in any temperature control device combination is greater than the preset update period number threshold, then the energy consumption prediction model is updated. The relaxed update strategy means that in situations with high energy consumption anomalies, the model will be immediately updated as soon as the data for any combination of temperature control devices reaches the update threshold, in order to quickly respond to the urgent need for model accuracy during periods of high energy consumption anomalies. The preset abnormal date percentage threshold refers to the critical value for distinguishing between "high energy consumption risk" and "medium-low energy consumption risk" dates.

[0042] If the proportion of days with abnormal energy consumption in the past 30 days exceeds the preset threshold for the proportion of abnormal days, it indicates that the current energy consumption abnormality is relatively common, and the model needs to be updated frequently to accurately reflect the equipment behavior under high energy consumption conditions; at this time, the number of updates for the adjustment period of any temperature control equipment combination reaches the threshold, which immediately triggers the model update.

[0043] This step represents the scenario with the highest demand in the update analysis and processing strategy. Its significance lies in ensuring that the model always reflects the latest energy consumption patterns when energy consumption is abnormally high, thereby improving the accuracy of energy consumption prediction during high-risk periods by using the lowest update trigger threshold.

[0044] Case 3: If the proportion of abnormal energy consumption days is not above the preset threshold for the proportion of abnormal days, then the update processing strategy of the energy consumption prediction model is determined to be a filtering update strategy. When the number of updates for the adjustment period in any temperature control device combination is greater than the preset threshold for the number of update periods, and there are updates for the adjustment period in the target number of temperature control device combinations, then the energy consumption prediction model is updated.

[0045] The aforementioned screening and update strategy requires that, when the energy consumption anomaly level is moderate, at the same time any combination of equipment triggers the basic conditions, a target number of equipment combinations must also have updates during the adjustment period, in order to balance the timeliness of updates and the sufficiency of data. The target number of equipment refers to the minimum number of temperature regulation equipment combinations that must simultaneously have updates during the adjustment period when the screening and update strategy is triggered.

[0046] If the percentage of abnormal energy consumption days is between zero and the preset threshold for the percentage of abnormal days, it indicates that the abnormal energy consumption situation has been manifested but has not yet reached a high-risk level. At this time, the number of adjustment periods for any combination of equipment needs to exceed the threshold, and the number of target equipment combinations also needs to have updated adjustment periods. Only after both conditions are met will the model update be triggered.

[0047] This step represents an intermediate state between conservative and lenient update strategies. Its significance lies in balancing the timeliness of updates with the sufficiency of data under the background of moderate energy consumption risk, avoiding updating the global model based solely on local data from a single device combination, thereby ensuring the reliability of model updates.

[0048] S2 determines the combination of devices of interest in the temperature regulation device combination based on the update analysis and processing strategy and combined with the activation data of the temperature regulation device in different time periods. It is understood that the activation data of the temperature control device in different time periods is determined based on the combination of temperature control devices activated in different time periods.

[0049] Specifically, the temperature control equipment assembly is constructed by freely combining temperature control equipment in the pharmaceutical warehouse.

[0050] The so-called free combination refers to constructing an idempotent set of all temperature control devices in the pharmaceutical warehouse. That is, any combination of one or more temperature control devices can constitute a temperature control device combination, so as to cover all possible equipment combinations in the actual operation of the pharmaceutical warehouse.

[0051] Assuming the pharmaceutical warehouse has three temperature control devices: refrigeration unit A, refrigeration unit B, and dehumidifier C, then all possible combinations of temperature control devices include {A}, {B}, {C}, {A,B}, {A,C}, {B,C}, and {A,B,C}, totaling 7 types. Combinations that have actually occurred in operation are all considered valid temperature control device combinations and included in subsequent analysis.

[0052] This step clarifies the construction method of the temperature control equipment combination. Its significance lies in exhaustively listing all possible equipment combinations to ensure the completeness of the analysis and avoid missing the energy consumption patterns of some equipment combinations within a specific temperature range.

[0053] It is understandable that, such as Figure 3 As shown, the method for determining the combination of devices of interest in the temperature control device assembly is as follows: In this embodiment, based on the activation data of temperature regulation device combinations at different time periods, the matching ambient temperature ranges corresponding to different temperature regulation device combinations are determined. Using the matching ambient temperature ranges corresponding to different temperature regulation device combinations and the update analysis and processing strategy, the degree of adaptation of the temperature regulation data update of different temperature regulation device combinations under the current update analysis and processing strategy is determined. Based on the degree of adaptation of the temperature regulation data update of different temperature regulation device combinations, the device combinations to be focused on are determined, that is, the temperature regulation device combinations that are forcibly activated within certain outdoor ambient temperature ranges are determined, thereby further improving the efficiency and comprehensiveness of the energy consumption prediction model update processing.

[0054] S21 uses the activation data of the temperature regulating devices in different time periods to determine the time periods during which the temperature regulating devices are activated, and uses the time periods during which the temperature regulating devices are activated as the matching regulation time periods of the temperature regulating devices. The matching adjustment period refers to the specific set of time periods in the history of a certain temperature regulation device combination that is actually in the on state and performs temperature regulation. Each matching adjustment period corresponds to a record containing the combination identifier, the start and end times of the period, and the corresponding outdoor ambient temperature.

[0055] If the combination {A,C} (refrigeration unit A + dehumidifier C) is in the on state in multiple historical periods in the historical data, then all these periods are included in the matching adjustment period set of combination {A,C}; if combination {B,C} only appears in a few periods, then the number of its matching adjustment period sets is relatively small.

[0056] This step establishes a historical runtime profile for each temperature control device combination. Its significance lies in classifying and summarizing the activation data of all devices according to the combination dimension, providing a data foundation for subsequent statistics on matching ambient temperature ranges by combination index.

[0057] S22 determines the matching ambient temperature range corresponding to the temperature control device combination based on the matching adjustment period of the temperature control device combination; The matched ambient temperature range refers to the outdoor ambient temperature range in which the number of matched adjustment periods of a certain temperature control device combination is greater than the preset threshold for the number of adjustment periods. In other words, the device combination is frequently used within this outdoor temperature range, indicating that the device combination has a high degree of matching with the temperature control requirements of this temperature range.

[0058] Assuming the outdoor ambient temperature is divided into multiple continuous temperature ranges, for the combination {A,C}, the outdoor ambient temperature corresponding to all its matching adjustment periods is counted, and the count is performed separately for each temperature range. If the number of matching adjustment periods for the combination {A,C} exceeds the preset threshold for the number of adjustment periods within the temperature range [28℃, 35℃], then [28℃, 35℃] is determined as one of the matching ambient temperature ranges for the combination {A,C}.

[0059] This step establishes the correlation between equipment combinations and outdoor temperature ranges. Its significance lies in revealing the activation patterns of each equipment combination under different outdoor temperature conditions, providing a temperature-dimensional analytical basis for subsequently determining which equipment combinations have insufficient data accumulation under the current update strategy.

[0060] It is understood that the matching ambient temperature range corresponding to the temperature regulation device combination is the outdoor ambient temperature range where the number of matching regulation periods of the temperature regulation device combination is greater than the preset threshold for the number of regulation periods.

[0061] S23 determines the combination of devices of interest in the combination of temperature control devices based on the update analysis and processing strategy and the matching ambient temperature ranges corresponding to different combinations of temperature control devices.

[0062] The determination of the combination of devices of interest takes into account the requirements of the update analysis and processing strategy on the degree of data accumulation (conservative strategies require higher data accumulation, while lenient strategies require lower data accumulation) and the sufficiency of the historical frequency of each combination of devices in the matching ambient temperature range (measured by the adjustment matching coefficient).

[0063] This step combines the assessment of update strategy and the sufficiency of equipment combination data into the final determination of the equipment combination to focus on. Its significance lies in ensuring that the scope of the equipment combination to focus on is compatible with the current energy consumption risk level and data accumulation needs, avoiding frequent forced activation due to too many equipment combinations to focus on, which would interfere with normal temperature control data, and also avoiding long-term lack of model update data for some equipment combinations due to too few equipment combinations to focus on.

[0064] Specifically, based on the update analysis and processing strategy and the matching ambient temperature ranges corresponding to different combinations of temperature control devices, the combination of devices of interest in the temperature control device combination is determined, including: S231 determines the adjustment matching coefficient of the temperature control equipment combination by matching the ambient temperature range corresponding to different temperature control equipment combinations and combining the proportion of the number of days in the matching ambient temperature range. It then determines whether the average value of the adjustment matching coefficient of different temperature control equipment combinations is greater than the preset matching coefficient threshold. If so, it means that different temperature control equipment combinations can be used reliably. At this time, the update processing efficiency of the energy consumption prediction model is high, thereby realizing the reliable update of the energy consumption prediction model. Therefore, the combination of equipment of interest in the temperature control equipment combination is determined to be the temperature control equipment combination whose adjustment matching coefficient is not greater than the preset matching coefficient threshold. If not, proceed to step S232. The adjustment matching coefficient refers to the ratio of the number of historical dates corresponding to the matched ambient temperature range of a certain temperature control equipment combination to the total number of dates within the statistical period. The larger the value, the wider the outdoor temperature coverage and the more sufficient the data accumulation of the equipment combination in history. The number of dates with a matched ambient temperature range refers to the total number of days within the statistical period where the outdoor ambient temperature falls within the matched ambient temperature range of a certain equipment combination.

[0065] Assuming the statistical period is the past 60 days, for temperature control equipment combination {A,B}, its matched ambient temperature range is [32℃, 38℃]. The number of days within the 60 days where the outdoor temperature is [32℃, 38℃] is counted as a certain value. The ratio of this value to the 60 days is the control matching coefficient for combination {A,B}. The control matching coefficient is calculated for all equipment combinations, and then the average is taken. If the average value is greater than the preset matching coefficient threshold, it indicates that all equipment combinations have sufficient historical data accumulation, and only a few equipment combinations with control matching coefficients not greater than the preset matching coefficient threshold need to be included in the focus equipment combinations. If the average value is not greater than the preset matching coefficient threshold, it indicates that the overall data accumulation is poor, and further analysis proceeds to S232.

[0066] This step is the first level of judgment in determining the equipment combination to focus on. Its significance lies in the fact that when the data accumulation of the overall equipment combination is good, only a few combinations with low compatibility are listed as the objects of attention, so as to avoid excessive intervention in normal temperature control operation, while ensuring that equipment combinations with missing key data are supplemented in a timely manner.

[0067] S232 defines a combination of temperature-regulating devices with an adjustment matching coefficient not greater than a preset matching coefficient threshold as a poorly adapted combination. Based on the proportion of the poorly adapted combination in the temperature-regulating device combination and the adjustment matching coefficient of different temperature-regulating device combinations, the update adaptation coefficient of the energy consumption prediction model is determined. It is then determined whether the update adaptation coefficient of the energy consumption prediction model is greater than the preset adaptation coefficient threshold. If yes, proceed to step S233. If no, determine the combination of devices of interest in the temperature-regulating device combination as a temperature-regulating device combination with an adjustment matching coefficient not greater than a second preset matching coefficient threshold (greater than the preset matching coefficient threshold). The poorly adapted device combination refers to a temperature control device combination whose adjustment matching coefficient is not greater than a preset matching coefficient threshold, indicating that it has a low historical usage frequency and insufficient data accumulation. The update matching coefficient is a comprehensive indicator calculated by combining the proportion of poorly adapted device combinations with the adjustment matching coefficients of each device combination, measuring the sufficiency of the overall model's updated data. The second preset matching coefficient threshold is a second-level threshold value greater than the preset matching coefficient threshold, used to expand the scope of device combinations of interest when the update matching coefficient is low.

[0068] Assume that the update adaptation coefficient is calculated by multiplying (1 - the proportion of poorly adapted device combinations) by the sum of the mean deviations of each adjustment matching coefficient. If the coefficient is greater than the preset adaptation coefficient threshold, it means that the overall adaptation deviation is still within an acceptable range, and the process proceeds to S233 to make a final judgment based on the update strategy type. If the coefficient is not greater than the threshold, it means that the overall adaptation is poor, and the scope of the device combinations to be monitored needs to be expanded. All device combinations whose adjustment matching coefficient is not greater than the second preset matching coefficient threshold (which is higher than the preset matching coefficient threshold) are included in the device combinations to be monitored.

[0069] This step is the second layer of judgment in determining the focus of the device combination. Its significance lies in determining whether it is necessary to expand the scope of attention when the overall data accumulation of the device combination is insufficient, by updating the adaptation coefficient as a comprehensive indicator. It also distinguishes between two handling paths: "needing to combine strategy type judgment" and "directly expanding the scope of attention", making the identification of the focus of the device combination more refined and targeted.

[0070] S233 determines whether the update analysis and processing strategy belongs to a relaxed update strategy. If yes, it determines that the focus device combination in the temperature regulation device combination is a temperature regulation device combination whose adjustment matching coefficient is not greater than a preset matching coefficient threshold. If no, it determines that the focus device combination in the temperature regulation device combination is a temperature regulation device combination whose adjustment matching coefficient is not greater than a second preset matching coefficient threshold (greater than the preset matching coefficient threshold).

[0071] The aforementioned lenient update strategy has a high update timeliness. When the update compatibility coefficient is still within an acceptable range but the overall data accumulation is insufficient, the screening threshold for the combination of devices of interest should be high (based on the preset matching coefficient threshold, only the combination of devices with the lowest compatibility is included). On the other hand, if a conservative update strategy or a screening update strategy is adopted, the update timeliness is low, and the urgency of data accumulation is relatively high. The range of combinations of devices of interest should be appropriately expanded (using the second preset matching coefficient threshold as the threshold) to cover more poorly adapted combinations.

[0072] If the update adaptation coefficient is greater than the preset adaptation coefficient threshold, after entering S233, it is determined that the current update analysis and processing strategy is a lenient update strategy. Then, the combination of devices to be focused on is the combination of devices whose matching coefficient is adjusted to not be greater than the preset matching coefficient threshold, which has a relatively narrow range. If the update analysis and processing strategy is a screening update strategy, then the combination of devices to be focused on is the combination of devices whose matching coefficient is adjusted to not be greater than the second preset matching coefficient threshold (i.e., a more lenient threshold), which includes a larger number of combinations.

[0073] This step is the final branch for determining the portfolio of devices of interest. Its significance lies in incorporating the timeliness of updates corresponding to the update strategy into the scope decision of the portfolio of devices of interest.

[0074] It should be noted that the second preset matching coefficient threshold is greater than the preset matching coefficient threshold, that is, the matching degree requirement corresponding to the second preset matching coefficient threshold is lower than the preset matching coefficient threshold, so as to include more device combinations with moderate compatibility when expanding the range of devices to be monitored.

[0075] This embodiment completes the entire process from temperature control device activation data to identification of the device combination of interest through progressive analysis from S21 to S233. Its core value lies in the following aspects: First, by adjusting the matching coefficient, the data accumulation degree of the device combination under historical temperature conditions is quantified into a comparable single indicator, making the identification of the device combination of interest objective and reproducible. Second, through a three-layer progressive judgment (S231 overall mean → S232 update adaptation coefficient → S233 strategy type), the decision-making process for the scope of the device combination of interest is decomposed into a progressive screening process from the whole to the individual, ensuring that each layer of judgment is based on sufficient information to make a reasonable decision. Third, by introducing the type of update analysis and processing strategy into the final judgment of S233, the screening results of the device combination of interest are adaptively adjusted according to changes in the degree of energy consumption risk, realizing the dynamism and flexibility of the management of the device combination of interest.

[0076] Taking a pharmaceutical warehouse as an example, the warehouse is equipped with four temperature control devices: refrigeration unit A (rated power 15kW), refrigeration unit B (rated power 10kW), dehumidifier C (rated power 3kW), and fresh air unit D (rated power 2kW). The sampling period was set to 30 minutes, and energy consumption data was collected for a total of 2,880 time periods over nearly 60 days.

[0077] Energy consumption anomaly period and date identification stage: Energy consumption data for each time period has been recorded. The preset energy consumption threshold is 22 kWh / time period. Statistics show that there were 432 time periods with energy consumption exceeding 22 kWh, which were marked as abnormal energy consumption periods.

[0078] Phase S11: Calculate the percentage of abnormal energy consumption periods on a daily basis. Taking day 15 as an example, there are a total of 48 time periods (24 hours × 2 time periods / hour), of which 14 time periods have energy consumption exceeding 22kWh. The percentage of abnormal time periods = 14 ÷ 48 ≈ 0.292.

[0079] Phase S12: The preset duration percentage threshold is 0.25. If the percentage of day 15 (0.292) is greater than 0.25, day 15 is marked as an energy consumption anomaly day. After statistical analysis of all 60 days, a total of 22 days were marked as energy consumption anomaly days.

[0080] Phase S13: Percentage of days with abnormal energy consumption = 22 ÷ 60 ≈ 0.367. The preset threshold for the percentage of abnormal days is 0.30. Since 0.367 ≥ 0.30, it falls under Case 2, and the update analysis and processing strategy is determined to be a lenient update strategy: when the number of update periods for any combination of temperature control equipment exceeds the preset threshold for the number of update periods (set to 20 periods), the energy consumption prediction model will be updated.

[0081] Focus on the device combination identification stage: The actual temperature control equipment combinations in the pharmaceutical warehouse were statistically analyzed as follows: {A} (single refrigeration unit A), {A,B} (dual refrigeration units), {A,C} (refrigeration unit A + dehumidifier C), {A,B,C} (dual refrigeration units + dehumidifier), and {A,B,C,D} (all equipment combinations), for a total of 5 combinations.

[0082] Phase S21: Extract the matching adjustment period for each combination from 2880 time period data.

[0083] Combination {A}: 320 matching adjustment periods in total; Combination {A,B}: 480 matching adjustment periods in total; Combination {A,C}: 260 matching adjustment periods in total; Combination {A,B,C}: 380 matching adjustment periods in total; Combinations {A,B,C,D}: 85 matching adjustment periods in total; Phase S22: The outdoor ambient temperature is divided into 5 intervals: [10℃, 18℃), [18℃, 25℃), [25℃, 32℃), [32℃, 38℃), and [38℃, 45℃]. The preset threshold for the number of adjustment periods is 50.

[0084] Count the number of matching adjustment periods for each combination in each temperature range: Combination {A}: 180 items in [25℃, 32℃), 110 items in [32℃, 38℃), and less than 50 items in other ranges → Matching ambient temperature ranges: [25℃, 32℃), [32℃, 38℃); Combination {A,B}: 250 within [32℃, 38℃), 180 within [38℃, 45℃], and less than 50 for the rest → Matching ambient temperature ranges: [32℃, 38℃), [38℃, 45℃]; Combinations {A,C}: 160 within [25℃,32℃) and 90 within [32℃,38℃) → Matching ambient temperature ranges: [25℃,32℃) and [32℃,38℃); Combinations {A,B,C}: 200 within [32℃,38℃) and 160 within [38℃,45℃] → Matching ambient temperature ranges: [32℃,38℃) and [38℃,45℃]. Combinations {A,B,C,D}: 55 within [32℃,38℃), and fewer than 50 elsewhere → Matching ambient temperature range: [32℃,38℃); Phase S231: Calculate the adjustment matching coefficient for each combination. Count the number of days within 60 days where the outdoor temperature falls within the matched ambient temperature range for each combination: Combination {A} (matching intervals [25℃, 32℃) and [32℃, 38℃)): There are 42 days out of 60 days where the outdoor temperature falls within the two intervals → Adjust the matching coefficient = 42 ÷ 60 = 0.700; Combination {A,B} (matching intervals [32℃,38℃) and [38℃,45℃]): 38 days → Adjust matching coefficient = 38 ÷ 60 ≈ 0.633; Combination {A,C} (matching intervals [25℃,32℃) and [32℃,38℃)): 42 days → Adjust matching coefficient = 42 ÷ 60 = 0.700; Combinations {A,B,C} (matching intervals [32℃,38℃) and [38℃,45℃]): 38 days → Adjust matching coefficient = 38 ÷ 60 ≈ 0.633; Combination {A,B,C,D} (matching interval [32℃,38℃)): 28 days → Adjust matching coefficient = 28 ÷ 60 ≈ 0.467; The average value of the adjustment matching coefficient for each combination = (0.700 + 0.633 + 0.700 + 0.633 + 0.467) ÷ 5 = 3.133 ÷ 5 = 0.627.

[0085] The preset matching coefficient threshold is 0.60. If the average value of 0.627 > 0.60, the judgment result is "yes", then the combination of devices of interest is determined to be a temperature control device combination with a matching coefficient not greater than the preset matching coefficient threshold of 0.60.

[0086] The adjustment matching coefficient for the combination {A,B,C,D} is 0.467 ≤ 0.60, therefore the combination of devices of interest is determined to be {A,B,C,D}, which is a total of 1 combination of devices of interest.

[0087] S3 uses the activation data of the monitored device combination in different time periods to determine the matching ambient temperature range corresponding to the monitored device combination. Based on the degree of overlap of the matching ambient temperature ranges corresponding to different monitored device combinations and the number of monitored device combinations, the adjustment and processing strategy of the monitored device combination in different outdoor ambient temperature ranges is determined. Specifically, such as Figure 4 As shown, the method for determining the adjustment and processing strategy of the monitoring device combination in different outdoor ambient temperature ranges is as follows: In this embodiment, based on the degree of overlap of the matching ambient temperature ranges corresponding to different combinations of devices of interest and the number of such combinations, the impact of forcibly activating the devices of interest in different outdoor ambient temperature ranges on the timeliness of updating energy consumption data during the temperature regulation process of different temperature regulation devices and the devices of interest is determined. Based on the degree of impact on the timeliness of updating energy consumption data during the temperature regulation process of different temperature regulation devices and the devices of interest, the adjustment and processing strategy of the devices of interest in different outdoor ambient temperature ranges is determined. This further improves the reliability of updating and processing energy consumption data of the devices of interest during the temperature regulation process by reducing the impact of temperature regulation devices excluding the devices of interest.

[0088] S31 determines the outdoor ambient temperature range belonging to the matching ambient temperature range based on the degree of overlap of the matching ambient temperature ranges corresponding to different combinations of concerned devices, and takes the outdoor ambient temperature range belonging to the matching ambient temperature range as the associated device combination of the outdoor ambient temperature range. The associated device combination refers to the set of devices of interest that are frequently used within a certain outdoor ambient temperature range, which is a temperature range that matches the ambient temperature range.

[0089] Assuming the device combination of interest is {A,B,C,D}, and its matching ambient temperature range is [32℃,38℃); then for the outdoor ambient temperature range [32℃,38℃), its associated device combination is {{A,B,C,D}}, which contains 1 device combination of interest; for the temperature range [25℃,32℃), since the matching ambient temperature range of {A,B,C,D} does not include this range, the associated device combination for this range is an empty set, containing 0 device combinations of interest.

[0090] This step establishes a mapping between outdoor temperature ranges and the natural activation relationship of the monitored equipment combination. Its significance lies in distinguishing between the two states of the monitored equipment combination within each temperature range: "naturally accumulating data" and "requiring mandatory intervention to accumulate data," thus providing a basis for the subsequent calculation of the regulation impact value.

[0091] S32 determines the regulation impact value within the outdoor ambient temperature range by summing the number of associated device combinations within the outdoor ambient temperature range with the number of temperature regulation device combinations belonging to the matching ambient temperature range. The number of associated device combinations within the outdoor ambient temperature range refers to the number of device combinations of interest that belong to the matched ambient temperature range within that temperature range. The number of temperature regulation device combinations belonging to the matched ambient temperature range refers to the total number of temperature regulation device combinations (including non-interested device combinations) within that temperature range. The larger this number, the more device combinations are naturally operating within that temperature range, and the greater the interference to the normal data accumulation of other device combinations when forcibly activating the interested device combinations.

[0092] The adjustment impact value within the outdoor ambient temperature range = (number of associated device combinations + number of temperature regulation device combinations belonging to the matched ambient temperature range) × preset scaling factor.

[0093] Assuming the preset scaling factor is 0.50. For the temperature range [32℃, 38℃): the number of associated device combinations is 1 (combination {A, B, C, D}), and the total number of temperature regulating device combinations belonging to this temperature range is 4 ({A}, {A, B}, {A, C}, and {A, B, C} all include this range), then the regulation impact value = (1 + 4) × 0.50 = 2.50. For the temperature range [25℃, 32℃): the number of associated device combinations is 0, and the total number of temperature regulating device combinations belonging to this range is 2 ({A}, {A, C}), then the regulation impact value = (0 + 2) × 0.50 = 1.00.

[0094] This step quantifies the activity level of the equipment combination within the outdoor ambient temperature range into a regulation impact value. Its significance lies in assessing, in a uniform numerical form, the impact of forcibly activating the monitored equipment combination on the acquisition of normal operation data of other equipment combinations within this temperature range, providing a quantitative basis for the differentiated configuration of subsequent regulation and processing strategies across different temperature ranges.

[0095] It should be noted that the adjustment impact value within the outdoor ambient temperature range is determined by multiplying the sum of the number of associated device combinations within the outdoor ambient temperature range and the number of temperature regulation device combinations belonging to the matching ambient temperature range by a preset proportional factor. The larger the adjustment impact value, the greater the impact on the efficiency of acquiring normal temperature regulation data of other temperature regulation device combinations when the device combination under attention is forcibly turned on and performs temperature regulation processing within the outdoor ambient temperature range.

[0096] S33 determines the adjustment and processing strategy of the combination of devices of interest in different outdoor ambient temperature ranges based on the adjustment impact value in different outdoor ambient temperature ranges and the number of the combination of devices of interest.

[0097] The number of device combinations of interest refers to the total number of temperature control device combinations currently identified as such. The preset threshold for the number of device combinations of interest refers to the critical value that distinguishes whether the number of device combinations of interest has a significant impact on the control strategy.

[0098] Assuming there is only one device combination to focus on (combination {A,B,C,D}), and if the preset threshold for the number of device combinations to focus on is 2, then 1 < 2, which is a case where the number of device combinations to focus on is small. The degree of interference with the normal operation of other device combinations is relatively controllable, and a forced activation strategy can be configured directly according to simplified rules. If the number of device combinations to focus on reaches 3, exceeding the threshold, then a strategy needs to be configured by analyzing the adjustment impact value interval by interval.

[0099] This step is the overall entry point for adjusting the processing strategy. Its significance lies in determining the level of granularity of the strategy configuration by focusing on the number of device combinations. When the number is small, the strategy configuration is simple and efficient, while when the number is large, it is necessary to start interval-by-interval fine-grained judgment, thereby achieving a reasonable balance between decision complexity and strategy accuracy.

[0100] It should be noted that, based on the adjustment impact values ​​within different outdoor ambient temperature ranges and the number of the devices under interest, the adjustment and processing strategies for the devices under interest within different outdoor ambient temperature ranges are determined, specifically including: Case 1: If the number of the monitored device combinations is less than the preset threshold for the number of monitored device combinations, then the forced use of the monitored device combinations in different outdoor ambient temperature ranges has little impact on the energy consumption data acquisition efficiency of other temperature regulation device combinations during the temperature regulation process. Therefore, the regulation processing strategy for the monitored device combinations in different outdoor ambient temperature ranges is determined to be to force activation processing in the matching ambient temperature range that does not belong to the monitored device combinations according to the preset number of temperature regulation time periods. That is, when temperature regulation is required, the monitored device combinations are used for temperature regulation processing after each preset number of temperature regulation time periods, i.e., in the next temperature regulation time period. The preset number of temperature adjustment periods refers to the periodic interval at which the focus device combination is forcibly activated. That is, after each of these number of temperature adjustment periods, the focus device combination is forcibly activated to perform a temperature adjustment process once in the next temperature adjustment period, so as to periodically accumulate the energy consumption data of the combination in the mismatched temperature range.

[0101] Assuming the number of monitored device combinations is 1 (less than the preset threshold of 2), then within the ranges of [10℃, 18℃), [18℃, 25℃), and [25℃, 32℃) (these ranges do not belong to the matching ambient temperature range of {A, B, C, D} [32℃, 38℃)), after each preset number of temperature adjustment periods (e.g., 5 periods), {A, B, C, D} will be activated for temperature adjustment once in the next adjustment period to accumulate the energy consumption data of this combination in the low temperature range.

[0102] This step is a simplified strategy for when the number of devices is small. Its significance lies in accumulating data synchronously across all non-matching temperature ranges by uniformly and periodically forcing them to turn on, without the need for differentiated configuration for each range, thus reducing the complexity of strategy management.

[0103] Case 2: S331 If the number of the monitored device combinations is not less than the preset threshold for the number of monitored device combinations, then obtain the number of associated device combinations in the outdoor ambient temperature range, and determine whether the number of associated device combinations in the outdoor ambient temperature range is less than the threshold for the number of monitored device combinations. If yes, then determine that the adjustment processing strategy of the monitored device combination in the outdoor ambient temperature range is as follows: If the outdoor ambient temperature range does not belong to the matching ambient temperature range of the monitored device combination, then perform forced activation processing according to the preset number of temperature adjustment periods. That is, when temperature adjustment is required, after each preset number of temperature adjustment periods, the monitored device combination is used for temperature adjustment processing in the next temperature adjustment period. If not, proceed to step S332. The threshold for the number of monitored device combinations refers to the critical value used in scenario 2 to determine whether the number of monitored device combinations that are naturally activated within a certain outdoor ambient temperature range is sufficient.

[0104] Assuming the number of monitored device combinations is 3 (not less than the preset threshold of 2 for the number of monitored device combinations), the number of associated device combinations in each outdoor ambient temperature range is determined interval by interval. For a temperature range with 1 associated device combination, if 1 is less than the threshold of 2 for the number of monitored device combinations, the forced activation strategy for that range is directly determined. For a temperature range with 3 associated device combinations, 3 is not less than the threshold of 2, indicating that a large number of monitored device combinations have already been activated naturally in that range, and the process continues to step S332 to analyze whether forced activation is still necessary.

[0105] This step performs the first-level interval screening for Case 2. Its significance lies in directly adopting the standard forced activation strategy for temperature intervals with fewer associated device combinations, and only performing further impact value judgments for temperature intervals with more associated device combinations, thereby reducing unnecessary fine-grained analysis.

[0106] S332 takes the outdoor ambient temperature range where the focus device combination is forcibly turned on as the filtering range, and determines whether the number of the filtering ranges is greater than the preset filtering range number threshold. If so, it is determined that the adjustment processing strategy of the focus device combination in the outdoor ambient temperature range excluding the filtering range does not need to be forcibly turned on. If not, the adjustment processing strategy of the focus device combination in the outdoor ambient temperature range is determined according to the adjustment influence value in the outdoor ambient temperature range. The selected temperature range refers to the set of temperature ranges within the entire outdoor ambient temperature range that have been determined to require forced activation of the combination of devices of interest. The preset threshold for the number of selected temperature ranges refers to the upper limit on the total number of temperature ranges that can be forced to activate. Once this threshold is exceeded, no new temperature ranges will be added to avoid severely interfering with the accumulation of normal temperature control data due to excessive forced activation.

[0107] Assuming the preset threshold for the number of filter intervals is 2, and 3 temperature intervals have already been identified as filter intervals; if it is necessary to determine whether to add a forced opening strategy for the 4th temperature interval, since the number of filter intervals has reached the threshold of 3, it is directly determined that this temperature interval does not need to be forced to open; if the number of filter intervals is only 2 (not greater than the threshold), then it is necessary to continue to determine whether to add a forced opening strategy based on the adjustment impact value of this temperature interval.

[0108] This step sets a total upper limit for the forced activation range. Its purpose is to avoid large-scale interference with the natural operating data of each equipment combination by performing forced activation in too many temperature ranges, ensuring the quality of normal temperature control data while keeping the side effects of forced activation within an acceptable range.

[0109] It is understandable that the adjustment and processing strategy for the combination of devices of interest within the outdoor ambient temperature range is determined based on the adjustment impact value within the outdoor ambient temperature range, specifically including: If the adjustment impact value within the outdoor ambient temperature range is greater than a preset impact threshold, then the adjustment processing strategy for the monitored device combination within the outdoor ambient temperature range is determined to be that no forced activation is required. If not, then the adjustment processing strategy for the monitored device combination within the outdoor ambient temperature range is determined to be that if the outdoor ambient temperature range does not belong to the matching ambient temperature range of the monitored device combination, then forced activation is performed according to a preset number of temperature adjustment periods. That is, when temperature adjustment is required, after each preset number of temperature adjustment periods, the monitored device combination is used for temperature adjustment processing in the next temperature adjustment period.

[0110] The preset influence threshold refers to the critical adjustment influence value that distinguishes whether the interference level of forced opening behavior within a certain temperature range exceeds the acceptable range.

[0111] Assuming the preset impact threshold is 2.00, and the adjustment impact value for a certain temperature range is 2.50 (greater than 2.00), it indicates that forcibly activating the monitored device combination in this range would significantly interfere with the normal data acquisition of other device combinations. Therefore, forced activation is not performed in this range. The adjustment impact value for another temperature range is 1.00 (less than 2.00), which is a lower level of interference. Therefore, the forced activation strategy with a preset cycle is executed.

[0112] This step is a fine-grained judgment at the interval level when the total amount of the screening interval has not reached the upper limit. Its significance lies in using the adjustment impact value as the final quantitative criterion for whether to perform forced activation, ensuring that forced activation only occurs within the temperature range where the degree of interference is controllable, thereby achieving a fine balance between the data accumulation of the focus equipment combination and the normal data quality of other equipment combinations.

[0113] This embodiment completes the entire process configuration from the focus on device combinations to the interval-based adjustment strategy through progressive analysis from S31 to S332. Its core value lies in the following aspects: First, by quantitatively calculating the impact value of the adjustment, the degree of interference of forced activation behavior on normal temperature control data is presented in numerical form, giving the strategy configuration a calculable objective basis; second, by hierarchically judging the number of focus device combinations, a simplified strategy is used to reduce management complexity when the number is small, while a refined interval-by-interval judgment is used to ensure strategy accuracy when the number is large; third, by using a constraint mechanism to screen the upper limit of the total number of intervals, the unlimited expansion of the forced activation range is prevented, ensuring that the overall temperature control data quality does not decline due to large-scale forced intervention.

[0114] S4 determines the combination of devices of interest for forced regulation within different outdoor ambient temperature ranges based on the regulation and processing strategy, and uses these as the combination of devices of interest. Based on the combination of devices of interest within different outdoor ambient temperature ranges and the degree of correlation with temperature regulation and processing data, the energy consumption analysis method of the energy consumption prediction model is determined.

[0115] The "affected equipment combination" refers to the equipment combination identified as requiring forced activation within a certain outdoor ambient temperature range, based on the adjustment strategy determined in S3. This means the equipment combination that influences the accumulation of temperature control data through forced intervention within that temperature range. The "temperature adjustment processing data" refers to the recorded data, such as energy consumption, adjustment duration, and adjustment effect, generated by the actual temperature adjustment operations performed by the temperature adjustment equipment combination within each outdoor ambient temperature range. The "correlation degree with temperature adjustment processing data" refers to the proportion of temperature adjustment processing data generated by the forced activation of the affected equipment combination within a certain outdoor ambient temperature range, and its impact on the representativeness of normal data within that range. The "energy consumption analysis method" refers to the analysis strategy that determines under what energy consumption prediction accuracy conditions and in which outdoor temperature ranges the forced activation of the equipment combination of interest is stopped, thereby reducing the additional energy consumption caused by forced activation while ensuring model reliability.

[0116] Assume that the equipment combination {A,B,C,D} is subject to forced activation in all three temperature ranges: [10℃,18℃), [18℃,25℃), and [25℃,32℃). Within the [25℃,32℃] range, due to the higher frequency of forced activation of {A,B,C,D}, the forced activation adjustment period generated by {A,B,C,D} accounts for a larger proportion of all adjustment periods in this range. This means it has a higher correlation with temperature regulation data, implying that disabling the forced activation strategy would significantly reduce energy consumption prediction data in this range, having a greater impact on model prediction accuracy. However, within the [10℃,18℃] range, because there are more natural adjustment periods, the forced activation affecting the equipment combination accounts for only a minority, and its correlation with temperature regulation data is lower. Therefore, stopping forced activation has a relatively smaller impact on the model.

[0117] This step combines the execution results of the adjustment processing strategy with the accuracy conditions of the energy consumption prediction model. Its significance lies in dynamically controlling the duration of the forced start behavior on the premise of achieving the accuracy target. After the model prediction is accurate enough, the forced start is stopped to reduce the additional energy consumption caused by human intervention, thus realizing a closed-loop optimization mechanism of "forced start to accumulate data - prediction accuracy improvement - gradual withdrawal of forced start".

[0118] Specifically, the method for determining the energy consumption analysis method of the energy consumption prediction model is as follows: In this embodiment, based on the combination of influencing devices in different outdoor temperature ranges and their correlation with temperature regulation processing data, the impact of frequent forced activation of influencing devices on energy consumption during temperature regulation in different outdoor temperature ranges is determined. Based on this impact, the energy consumption analysis method for the energy consumption prediction model is determined. Specifically, under what level of accuracy in the energy consumption prediction model's energy consumption forecast, the analysis of energy consumption during forced activation in certain outdoor temperature ranges will no longer be performed. This further reduces energy consumption during temperature regulation while ensuring the reliability of the energy consumption prediction model's update process.

[0119] S41 determines the number of influencing device combinations in different outdoor ambient temperature ranges based on the combination of influencing devices in different outdoor ambient temperature ranges, and takes the outdoor ambient temperature range in which the number of influencing device combinations is greater than the preset threshold for the number of influencing device combinations as the influencing temperature range. The number of affected device combinations refers to the number of device combinations identified as requiring forced activation within a certain outdoor ambient temperature range. The affected temperature range refers to the outdoor ambient temperature range where the number of affected device combinations exceeds a preset threshold, indicating that forced activation has a high impact on energy consumption data within this temperature range, and should be treated differently when deciding when to stop forced activation.

[0120] Assuming the preset threshold for the number of affected device combinations is 2, if 3 combinations of devices of interest are forcibly turned on within a certain temperature range, then 3 > 2, and this range is marked as an affected temperature range; if only 1 combination of devices of interest is forcibly turned on within another temperature range, then 1 ≤ 2, and it is not marked as an affected temperature range.

[0121] This step categorizes the outdoor temperature ranges according to the degree of impact of forced activation. Its significance lies in distinguishing between two situations: "requiring stricter stopping conditions" (affected temperature range) and "allowing for more lenient stopping" (non-affected temperature range), providing a classification basis for the differentiated configuration of subsequent energy consumption analysis methods.

[0122] The above steps include the following: Determine if there is an influencing temperature range. If yes, proceed to step S42. If no, determine that the energy consumption analysis method of the energy consumption prediction model is a rigorous analysis method. The aforementioned rigorous analysis method refers to stopping the forced opening of all temperature ranges only after the energy consumption prediction model meets stricter overall accuracy conditions, under the condition that there are no influencing temperature ranges (i.e., the forced opening impact is low in all temperature ranges).

[0123] It should be noted that the strict analysis method means that only when the average deviation rate of the energy consumption of the energy consumption prediction model from the energy consumption of the energy consumption prediction model in all temperature regulation periods in the period after the most recent update is less than a preset deviation rate threshold, and the number of temperature regulation periods in different outdoor temperature ranges is greater than the target period number threshold, is it not necessary to forcibly turn on the combination of devices under interest. That is, energy consumption analysis processing of the combination of devices under interest during the forced turn-on process is no longer performed. In this way, under the premise of high prediction reliability, the occurrence of excessive energy consumption caused by forced turn-on is minimized.

[0124] The deviation rate refers to the ratio of the absolute value of the difference between the actual energy consumption and the energy consumption predicted by the energy consumption prediction model during a certain temperature regulation period to the actual energy consumption, reflecting the relative error level of the model prediction. The target period number threshold refers to the minimum number of temperature regulation periods that need to be accumulated in each outdoor temperature range to ensure that the model has sufficient data support in each temperature range.

[0125] Assuming that the number of affected equipment combinations in all outdoor temperature ranges does not exceed the preset threshold for the number of affected equipment combinations, a rigorous analysis method is adopted. When the average deviation rate of all temperature regulation periods in the most recently updated period of the energy consumption prediction model is lower than the preset deviation rate threshold (e.g., 5%), and the number of regulation periods in all outdoor temperature ranges exceeds the target number of periods threshold (e.g., 30 periods), the forced opening process for all temperature ranges is stopped.

[0126] This step represents the optimal decision-making process for energy consumption analysis under the condition of minimal impact. Its significance lies in the fact that when the overall impact of forced activation is low, no interval-level differentiation is required. Instead, the overall accuracy is uniformly used to determine the timing of cessation, simplifying the management process while ensuring model accuracy.

[0127] S42 determines the temperature regulation period within the affected temperature range based on the degree of correlation with the temperature regulation processing data; The temperature adjustment period within the affected temperature range refers to the temperature adjustment period associated with the energy consumption data generated by the forced start-up of the affected equipment combination within that affected temperature range. The adjustment processing data during these periods is directly related to the forced start-up behavior of the affected equipment combination, and the higher its proportion, the greater the impact of forced start-up on the energy consumption data in that range.

[0128] Assuming the affected temperature range is [25℃, 32℃), count the number of adjustment periods caused by the forced activation of the affected equipment combination {A, B, C, D} in all adjustment periods within this range, and calculate its proportion in all adjustment periods in this range, as an indicator to measure the degree of correlation with temperature regulation processing data.

[0129] This step provides a data foundation for a detailed analysis of the temperature ranges influencing the data. Its significance lies in quantifying the impact of forced activation on the data structure of each affected temperature range by the degree of correlation with temperature regulation processing data, so that the subsequent judgment conditions for stopping forced activation can be set differently according to the degree of influence.

[0130] It should be noted that the above steps include the following: Obtain the number of the affected temperature ranges, and determine whether the number of the affected temperature ranges is greater than the preset threshold for the number of affected temperature ranges. If yes, determine that the energy consumption analysis method of the energy consumption prediction model is a classification analysis method; otherwise, proceed to step S43. The classification analysis method refers to an analysis strategy that, when there are a large number of affected temperature ranges, configures different stop-force-on conditions for affected temperature ranges and non-affected temperature ranges.

[0131] It is understood that the classification analysis method stipulates that if the outdoor temperature range belongs to the affected temperature range, then only when the average deviation rate between the energy consumption of the energy consumption prediction model and the energy consumption of the energy consumption prediction model in all temperature regulation periods in the period after the most recent update is less than a preset deviation rate threshold, and the number of temperature regulation periods in the outdoor temperature range is greater than the target period number threshold, will it be unnecessary to forcibly turn on the focus device combination, i.e., no energy consumption analysis processing will be performed on the focus device combination during the forced turn-on process; if the outdoor temperature range does not belong to the affected temperature range, then only when the average deviation rate between the energy consumption of the energy consumption prediction model and the energy consumption of the energy consumption prediction model in all temperature regulation periods in the period after the most recent update is less than a preset deviation rate threshold, and the number of temperature regulation periods in different outdoor temperature ranges is greater than the target period number threshold, will it be unnecessary to forcibly turn on the focus device combination, i.e., no energy consumption analysis processing will be performed on the focus device combination during the forced turn-on process.

[0132] In the classification analysis method, the temperature-affected interval uses the number of time periods within the interval as a separate condition (only the number of time periods in the interval is required to meet the standard), while the non-temperature-affected interval uses the number of time periods of all intervals as a whole condition (all outdoor temperature intervals are required to meet the standard), which reflects the difference in the threshold for stopping forced opening when the degree of impact is different.

[0133] Assuming the affected temperature range is [32℃, 38℃), the forced opening of this range can be stopped when the number of adjustment periods in this range exceeds the target number of time periods threshold and the average deviation rate of all ranges is lower than the preset deviation rate threshold. For non-affected temperature ranges (such as [10℃, 18℃)), the forced opening can only be stopped when the number of adjustment periods in all outdoor temperature ranges exceeds the target number of time periods threshold, which is a more stringent condition.

[0134] This step is a processing strategy when there are many temperature ranges affected. Its significance lies in setting classified stop conditions so that temperature ranges that are more affected by forced opening have a more lenient stop threshold (only this range needs to meet the standard), while ranges that are less affected have a more stringent overall compliance condition, thereby achieving a fine-grained match between accuracy requirements and the timing of forced opening exit.

[0135] S43 determines the energy consumption analysis method of the energy consumption prediction model based on the affected temperature range and the temperature adjustment period within different affected temperature ranges.

[0136] The final determination of the energy consumption analysis method integrates the analysis results of the number of temperature ranges affected and the degree of correlation between each temperature range and the temperature regulation processing data (quantified by the influence weight value).

[0137] If the number of affected temperature intervals does not exceed the preset threshold for the number of affected temperature intervals (the process proceeds to S43 instead of using a classification analysis method), then S43 will further analyze the correlation of adjustment periods within the affected temperature intervals, and determine whether to use a classification analysis method or a secondary classification analysis method based on whether the sum of the influence weight values ​​exceeds the preset weight threshold.

[0138] This step is a fine-grained decision-making path for the energy consumption analysis method when the number of affected temperature ranges is moderate. Its significance lies in the fact that when the number of affected temperature ranges is neither too many (not triggering S42 direct classification) nor zero (not triggering S41 strict analysis), the overall degree of influence is further identified through weighted analysis of the influence weight values, so as to ensure that the final selected energy consumption analysis method is accurately matched with the actual forced activation influence structure.

[0139] It is understood that the above steps include the following: Based on the proportion of temperature regulation periods within different temperature ranges, the influence weight value of the temperature range is determined. It is then determined whether the sum of the influence weight values ​​of different temperature ranges is greater than a preset weight threshold. If so, the frequent operation of the monitored equipment combination may lead to an increase in energy consumption within the aforementioned temperature range, and the energy consumption analysis method of the energy consumption prediction model is determined to be a classification analysis method. If not, the energy consumption analysis method of the energy consumption prediction model for the temperature range is determined to be a secondary classification analysis method.

[0140] The influence weight value of the temperature range refers to the ratio of the number of temperature adjustment periods within the temperature range to the total number of adjustment periods. The larger the value, the greater the impact of forced activation on overall energy consumption.

[0141] This step summarizes the overall impact of forced activation across all temperature ranges into a single criterion by summing the weighted values. The significance of this step is that when the overall impact of forced activation is high (and exceeds the threshold), the same classification analysis method as S42 is used to ensure consistency. When the impact is low, a more refined secondary classification analysis method is used to perform interval-level differentiation processing, ensuring that the method selection is highly matched with the actual impact level.

[0142] It should be noted that the aforementioned secondary classification analysis method specifically includes: If the influence weight value of the temperature range is above the preset influence weight threshold, the frequent activation of the device combination may lead to an increase in energy consumption within the temperature range. Therefore, only when the average deviation rate between the energy consumption of the energy prediction model and the energy consumption of the energy prediction model in all temperature adjustment periods after the most recent update is less than the preset deviation rate threshold, and the number of temperature adjustment periods within the temperature range is above the target period number threshold, will it be unnecessary to force the device combination to be activated, i.e., energy consumption analysis processing of the device combination during the forced activation process will no longer be performed. If the influence weight value of the temperature range is not above the preset influence weight threshold, then only if the average deviation rate of the energy consumption of the energy consumption prediction model from the energy consumption of the energy consumption prediction model in all temperature adjustment periods in the period after the most recent update is less than the preset deviation rate threshold, and the number of temperature adjustment periods in the outdoor temperature range including the preset number of ranges is above the target number threshold, then it is not necessary to force the combination of devices under concern to be turned on, that is, no longer perform energy consumption analysis processing on the combination of devices under concern during the forced turn-on process; If the temperature range is not within the affected temperature range, then only if the average deviation rate between the energy consumption of the energy consumption prediction model and the energy consumption of the energy consumption prediction model in all temperature regulation periods in the period after the most recent update is less than the preset deviation rate threshold, and the number of temperature regulation periods in different outdoor temperature ranges is greater than the target period number threshold, then it is not necessary to force the combination of devices under interest to be turned on, that is, energy consumption analysis and processing of the combination of devices under interest during the forced turn-on process will no longer be performed.

[0143] The preset influence weight threshold refers to the critical weight value used within the secondary classification analysis method to distinguish the degree of influence of a certain temperature range. The preset number of ranges refers to the number of outdoor temperature ranges that need to meet their data volume conditions when the temperature range with a lower influence weight value is no longer forcibly opened; its value is greater than 1 and less than the total number of all ranges.

[0144] This step achieves fine differentiation within the secondary classification analysis method. Its significance lies in further distinguishing between two strategies based on the degree of mandatory activation correlation of each affected temperature range: "high correlation - only this range needs to meet the standard to stop" and "low correlation - multiple ranges including this range need to meet the standard simultaneously to stop". This avoids applying the same stopping conditions to all affected temperature ranges in a one-size-fits-all manner, thereby ensuring that the energy consumption prediction accuracy of each range reaches an acceptable level before the mandatory activation is stopped through refined management.

[0145] It should be noted that when adjusting the temperature, the energy consumption prediction model is used to determine the energy consumption of different combinations of temperature control equipment under the current outdoor and indoor temperatures. The combination of temperature control equipment with the lowest energy consumption that meets the adjustment time threshold requirement is used for temperature control.

[0146] The adjustment time refers to the predicted time required for a combination of temperature control devices to adjust the indoor temperature from its current value to the target temperature range under current conditions. The time threshold requirement means that the adjustment time must not exceed the preset maximum allowable adjustment time limit. The lowest energy consumption refers to the combination with the lowest predicted energy consumption among all temperature control device combinations that meet the adjustment time threshold requirement.

[0147] Assuming the current outdoor temperature is 30℃ and the indoor temperature is 26℃, the target temperature range is [22℃, 24℃]. The energy consumption prediction model predicts an adjustment time of 45 minutes and energy consumption of 12kWh for combination {A}, an adjustment time of 28 minutes and energy consumption of 15kWh for combination {A,C}, and an adjustment time of 35 minutes and energy consumption of 18kWh for combination {A,B}. The preset maximum allowable adjustment time is 40 minutes. The combinations that meet the adjustment time requirement (≤40 minutes) are {A,C} (28 minutes) and {A} (45 minutes, which exceeds the requirement). Among the combinations that meet the time threshold requirement, the one with the lowest energy consumption is {A,C} (15kWh). Therefore, {A,C} is used for temperature regulation.

[0148] This step combines the output of the energy consumption prediction model with the actual temperature regulation decision. Its significance lies in ensuring that the temperature control effect meets the standard under the premise of ensuring the regulation time constraint, and selecting the equipment combination with the lowest energy consumption through the energy consumption prediction model. This directly transforms the model's predictive ability into the actual benefits of energy-saving operation, demonstrating the practical value of energy consumption prediction analysis methods in guiding temperature control decisions.

[0149] This embodiment, through progressive analysis from S41 to S43, completes the entire process of determining the influencing equipment combination and the energy consumption analysis method. Its core value lies in the following aspects: First, by identifying the number of influencing equipment combinations and the influencing temperature range, the impact of forced start-up behavior within different temperature ranges is quantified and classified, ensuring that the conditions for stopping forced start-up are based on sound reasoning. Second, through a three-tiered progressive decision-making system—rigorous analysis method, classification analysis method, and secondary classification analysis method—energy consumption analysis strategies are precisely matched to different influencing patterns, avoiding oversimplification or overcomplication in method selection. Third, by comparing the influence weight value with the preset influence weight threshold, differentiated stop condition configuration at the interval level is further implemented within the secondary classification analysis method, maximizing the energy-saving benefits brought by stopping forced start-up after achieving the required energy consumption prediction accuracy.

[0150] It should be noted that when adjusting the temperature, the energy consumption prediction model is used to determine the energy consumption of different combinations of temperature control equipment under the current outdoor and indoor temperatures. The combination of temperature control equipment with the lowest energy consumption that meets the adjustment time threshold requirement is used for temperature control.

[0151] Example 2 Secondly, the present invention provides a computer system comprising: a memory and a processor connected in communication, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the above-described method for predicting and analyzing energy consumption in a pharmaceutical warehouse when running the computer program.

[0152] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of apparatus, devices, and non-volatile computer storage media are basically similar to the method embodiments, so the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0153] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0154] The above description is merely one or more embodiments of this specification and is not intended to limit this specification. Various modifications and variations can be made to the one or more embodiments of this specification by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of one or more embodiments of this specification should be included within the scope of the claims of this specification.

Claims

1. A method for predicting and analyzing energy consumption in a pharmaceutical warehouse, characterized in that, Specifically, it includes: Based on the energy consumption data of the pharmaceutical warehouse, the distribution data of abnormal energy consumption periods in the pharmaceutical warehouse is determined, and based on the distribution data of the abnormal energy consumption periods, the update analysis and processing strategy of the energy consumption prediction model is determined. Based on the aforementioned update analysis and processing strategy, and combined with the activation data of temperature control devices in different time periods, the combination of devices of interest in the temperature control device combination is determined. Based on the activation data of the monitored device combination at different time periods, the matching ambient temperature range corresponding to the monitored device combination is determined. Based on the degree of overlap of the matching ambient temperature ranges corresponding to different monitored device combinations and the number of monitored device combinations, the adjustment and processing strategy of the monitored device combination in different outdoor ambient temperature ranges is determined. Based on the aforementioned regulation and processing strategy, the combination of devices of interest for forced regulation and processing in different outdoor ambient temperature ranges is determined, and these are taken as the combination of influencing devices. By considering the combination of influencing devices in different outdoor ambient temperature ranges and their correlation with temperature regulation and processing data, the impact of frequent forced activation of the combination of influencing devices on energy consumption in the temperature regulation process in different outdoor ambient temperature ranges is determined. The energy consumption analysis method of the energy consumption prediction model is then determined, and it is decided when to stop forced activation. The equipment combination in question is a temperature control equipment combination that is forced to operate within a certain range of outdoor ambient temperature. The method for determining the combination of devices of interest in the temperature control equipment combination is as follows: Based on the activation data of the temperature control devices in different time periods, the time periods during which the temperature control device combination is activated are determined, and the time periods during which the temperature control device combination is activated are used as the matching adjustment time periods for the temperature control device combination. Based on the matching adjustment period of the temperature control equipment combination, the matching ambient temperature range corresponding to the temperature control equipment combination is determined. Based on the update analysis and processing strategy and the matching ambient temperature ranges corresponding to different combinations of temperature control devices, the combination of devices of interest in the temperature control device combination is determined, specifically including: S231 determines the adjustment matching coefficient of the temperature control device combination by considering the matching ambient temperature range corresponding to different temperature control device combinations and the proportion of dates that exist in the matching ambient temperature range. It then determines whether the average adjustment matching coefficient of different temperature control device combinations is greater than a first preset matching coefficient threshold. If so, it determines that the device combination of interest in the temperature control device combination is a temperature control device combination whose adjustment matching coefficient is not greater than the first preset matching coefficient threshold. If not, it proceeds to step S232. S232 defines a combination of temperature-regulating devices with an adjustment matching coefficient not greater than a first preset matching coefficient threshold as a poorly adapted combination. Based on the proportion of the poorly adapted combination in the temperature-regulating device combination and the adjustment matching coefficients of different temperature-regulating device combinations, the update adaptation coefficient of the energy consumption prediction model is determined. It is then determined whether the update adaptation coefficient of the energy consumption prediction model is greater than a preset adaptation coefficient threshold. If yes, proceed to step S233. If no, determine the combination of devices of interest in the temperature-regulating device combination as a combination of temperature-regulating devices with an adjustment matching coefficient not greater than a second preset matching coefficient threshold, wherein the second preset matching coefficient threshold is greater than the first preset matching coefficient threshold. S233 determines whether the update analysis and processing strategy belongs to a relaxed update strategy. If yes, it determines that the focus device combination in the temperature regulation device combination is a temperature regulation device combination with an adjustment matching coefficient not greater than a first preset matching coefficient threshold. If no, it determines that the focus device combination in the temperature regulation device combination is a temperature regulation device combination with an adjustment matching coefficient not greater than a second preset matching coefficient threshold.

2. The method for predicting and analyzing energy consumption in a pharmaceutical warehouse as described in claim 1, characterized in that, The energy consumption data of the pharmaceutical warehouse includes the energy consumption during the temperature regulation process at different times.

3. The method for predicting and analyzing energy consumption in a pharmaceutical warehouse as described in claim 1, characterized in that, The abnormal energy consumption period of the pharmaceutical warehouse is determined based on the period when the energy consumption of the pharmaceutical warehouse exceeds a preset energy consumption threshold.

4. The method for predicting and analyzing energy consumption in a pharmaceutical warehouse as described in claim 1, characterized in that, The method for determining the update analysis and processing strategy of the energy consumption prediction model is as follows: Based on the distribution data of the abnormal energy consumption periods, the duration percentage of the abnormal energy consumption periods on different dates is determined. The dates with abnormal energy consumption are determined by the percentage of time periods with abnormal energy consumption on different dates. Based on the distribution data of the energy consumption anomaly dates, an update analysis and processing strategy for the energy consumption prediction model is determined.

5. The method for predicting and analyzing energy consumption in a pharmaceutical warehouse as described in claim 4, characterized in that, The dates with abnormal energy consumption are those dates whose duration of abnormal energy consumption periods is greater than a preset duration percentage threshold.

6. The method for predicting and analyzing energy consumption in a pharmaceutical warehouse as described in claim 4, characterized in that, Based on the distribution data of the energy consumption anomaly dates, the update analysis and processing strategy for the energy consumption prediction model is determined, specifically including: If there are no abnormal energy consumption dates, the update processing strategy of the energy consumption prediction model is determined to be a conservative update strategy. That is, when the number of updates for the adjustment period in at least the target number of temperature control equipment combinations is greater than the preset threshold for the number of update periods, the energy consumption prediction model is updated.

7. The method for predicting and analyzing energy consumption in a pharmaceutical warehouse as described in claim 1, characterized in that, The activation data of the temperature control device at different times is determined based on the combination of temperature control devices activated at different times.

8. The method for predicting and analyzing energy consumption in a pharmaceutical warehouse as described in claim 1, characterized in that, The method for determining the adjustment and processing strategy of the aforementioned equipment combination in different outdoor ambient temperature ranges is as follows: Based on the degree of overlap of the matching ambient temperature ranges corresponding to different combinations of monitored devices, the monitored device combinations whose outdoor ambient temperature ranges belong to the matching ambient temperature ranges are determined, and the monitored device combinations whose outdoor ambient temperature ranges belong to the matching ambient temperature ranges are regarded as the associated device combinations of the outdoor ambient temperature ranges. The regulation impact value within the outdoor ambient temperature range is determined by the sum of the number of associated device combinations within the outdoor ambient temperature range and the number of temperature regulation device combinations belonging to the matching ambient temperature range. Based on the adjustment impact values ​​within different outdoor ambient temperature ranges and the number of the devices of interest combination, the adjustment and processing strategies of the devices of interest combination within different outdoor ambient temperature ranges are determined.

9. A computer system, comprising: A memory and processor connected by communication, and a computer program stored in the memory and capable of running on the processor, characterized in that, when the processor runs the computer program, it executes a method for predicting and analyzing energy consumption in a pharmaceutical warehouse as described in any one of claims 1-8.

Images