A collaborative energy-saving optimization method and system based on multi-source data analysis and judgment

Abstract

The application discloses a kind of based on multi-source data analysis research and judgment's collaborative energy-saving optimization method and system, it is related to building energy management and building automation control technical field, the present application is by being mapped to space unit and its equipment service relationship with reservation scheduling data, traffic perception data, wireless access data, indoor environment data, building control data and equipment operation data are unified, and form regional operation image under the unified time benchmark, then the regional operation image is executed cross-source consistency analysis to generate occupancy confidence state, reservation deviation state and indoor air quality conflict state, to determine the regional mark of space unit, so that scene recognition no longer depends on fixed schedule, single access signal or low frequency feedback quantity, but is based on multi-source information to actual occupancy and environmental response are jointly determined, to reduce the regional misjudgment under the condition that reservation and real presence are not synchronized.

Description

Technical Field

[0001] This invention relates to the field of building energy management and building automation control technology, and in particular to a collaborative energy-saving optimization method and system based on multi-source data analysis and judgment. Background Technology

[0002] During daytime operation, the mixed-use office and teaching park exhibits a clear temporal coupling relationship between meeting reservations, course scheduling, access control, wireless access, indoor environmental changes, and the operational status of building equipment. Different spatial units may be in different operational states at the same time, such as continuous attendance, temporary gatherings, short-term vacancy, reservation deviations, or rapid switching, which further leads to differences in the demand for fresh air, air supply, thermal environment regulation, and related energy-consuming objects.

[0003] Existing collaborative energy-saving control methods are typically based on fixed-schedule control, single-occupancy signal-linked control, model predictive control, or reinforcement learning control. These methods use reservation information, access control information, occupancy information, or environmental feedback to coordinate and adjust subsystems such as air conditioning, ventilation, and lighting. During operation, these methods often rely primarily on preset time sequences, single-source status, or local feedback results as the main control basis. When the scheduled appointments are not synchronized with actual attendance, or when temporary meetings, extended sessions, class rescheduling, or short-term gatherings occur within the space unit, single-source status cannot reliably reflect the true occupancy changes, easily leading to lag or misjudgment in space unit status identification.

[0004] In the above situation, if the control system still controls according to a fixed schedule, a single access control result, or low-frequency environmental feedback, the vacant space unit may continue to maintain the guaranteed working condition, and the temporarily full-load space unit may still operate under the normal working condition. This will lead to the inconsistency between the fresh air, air supply and thermal environment adjustment and the actual demand, the equipment will switch back and forth between adjacent control modes, the ventilation response and adjustment response of the local space unit will lag, and the deviation between the control result and the on-site operating status will increase.

[0005] On the other hand, appointment scheduling data, access sensing data, wireless access data, indoor environmental data, building control data, and equipment operation data are usually scattered across different business and control systems, with differences in time granularity, state semantics, and spatial object mapping among the data sources. Even when multiple data sources are introduced into existing solutions, they often remain within a single controller or subsystem for localized processing, making it difficult to form unified data alignment results, state determination results, and control triggering results around the same spatial unit. When data conflicts, execution deviations, or manual intervention occur during operation, the control system struggles to identify the source of the conflict in a timely manner and maintain the current control state stably, further increasing the risk of misjudgment and instability in the control chain.

[0006] Therefore, how to uniformly align and consistently assess multi-source operational data around spatial units during the daytime operation of a mixed office and teaching park, and thereby stably form a continuous closed loop for regional status determination, control triggering, and anomaly rollback, has become a technical problem that needs to be solved. Summary of the Invention

[0007] This application provides a collaborative energy-saving optimization method and system based on multi-source data analysis and judgment, which solves the problem of scene misjudgment and control switching instability caused by multi-source mismatch during the day in mixed parks.

[0008] In a first aspect, embodiments of the present invention provide a collaborative energy-saving optimization method based on multi-source data analysis and judgment, applied to the daytime operation phase of a mixed-use office and teaching park, comprising: Step S1: Collect reservation scheduling data, access sensing data, wireless access data, indoor environment data, building control data, and equipment operation data, and map these data to spatial units and their equipment service relationships; Step S2: Align the mapped data according to a unified time benchmark, and combine spatial adjacency, equipment service relationship and business affiliation relationship to form a regional operation profile of the spatial unit; Step S3: Perform cross-source consistency analysis on the regional operation profile to generate occupancy confidence status, reservation deviation status and indoor air quality conflict status, and determine the regional label of the spatial unit; Step S4: Generate control boundaries and sequence holding conditions based on the area markers, equipment interlocking relationships, manual intervention holding status, and the current collaborative control sequence; upon first entering automatic control, write the default control sequence into the current collaborative control sequence; only after the minimum holding time is met and the prohibition of reverse switching is lifted, select a collaborative control sequence matching the control boundaries and sequence holding conditions from the control sequence library, and issue it to the target device according to the hierarchy of area-level controllers and equipment-level controllers; otherwise, maintain the current collaborative control sequence. Step S5: Collect execution feedback data and verify its consistency with the region marker. If any of the following conditions are met, switch to the conservative control sequence and update the region marker: cross-source conflict persists; execution deviation persists; manual intervention maintains the status.

[0009] In some embodiments, the appointment scheduling data includes meeting appointment data, course scheduling data, and temporary scheduling data; The access control data includes access control data and area-to-departure records; The wireless access data includes wireless terminal access status data; The indoor environmental data includes carbon dioxide data, temperature data, and humidity data; The building control data includes damper status data, valve status data, supply and return air status data, and setpoint data; The equipment operation data includes start / stop status data, operation mode data, and fault alarm data for air handling equipment, fresh air equipment, terminal control equipment, lighting equipment, and sunshade equipment.

[0010] In some embodiments, performing cross-source consistency analysis includes: Within the same analysis window, the appointment arrival deviation result is obtained based on the appointment scheduling data and access perception data; the occupancy change result is obtained based on the access perception data and wireless access data; the ventilation response result is obtained based on the occupancy change result, indoor environment data, building control data, and equipment operation data; and the occupancy confidence state, appointment deviation state, and indoor air quality conflict state are determined based on the appointment arrival deviation result, occupancy change result, and ventilation response result.

[0011] In some embodiments, the region markings include a protection region, an adjustment region, a downgrade region, and a region to be observed, and only one result is retained within the same analysis window; Different spatial types and different time periods call the area marking judgment rule set respectively. The area marking judgment rule set includes at least the attendance ratio judgment item, online change judgment item, carbon dioxide change judgment item, duration judgment item, consistency judgment item, and manual intervention maintenance duration judgment item. Among them, continuous attendance is determined by the joint result of appointment scheduling data, access perception data, and wireless access data; sudden increase in occupancy is determined by the joint result of the change in the number of entries, the change in the number of online terminals, and the carbon dioxide change trend in adjacent analysis windows; continuous low occupancy is determined by the results of low attendance, low online and no indoor air quality conflict in continuous analysis windows; and manual intervention maintenance status is determined by the joint result of manual issuance time, manual release status, and current controller lock status. When a spatial unit is characterized as continuously present within a preset continuous time period and the indoor air quality conflict status is characterized as having a conflict, it is marked as a protected area. When a space unit is characterized as continuously low occupancy and indoor air quality conflict status is characterized as non-conflict within a preset continuous time period, it is marked as a downgraded zone. A spatial unit will be marked as an adjustment zone under any of the following circumstances: the spatial unit is characterized by a sudden increase in occupancy within a preset continuous time period; or the reservation deviation status is characterized by the reservation arrival deviation exceeding a threshold. A spatial unit is marked as an observation area under any of the following conditions: the cross-source consistency of the spatial unit is below the consistency threshold; or manual intervention maintains the status. When a conflict occurs in the region label, the final region label is determined in the order of the observation region, the protection region, the adjustment region, and the downgrade region. When the region label changes within the same spatial unit in a continuous analysis window, the sequence preservation condition is verified first, and then the region label is updated.

[0012] In some embodiments, the control boundaries include a fresh air lower limit boundary, an air supply regulation boundary, a setpoint adjustment boundary, and a lighting and shading linkage boundary; the sequence holding conditions include a minimum holding time, a condition prohibiting reverse switching, and a condition requiring manual intervention. The fresh air lower limit boundary corresponds to the lowest fresh air capacity level that a spatial unit is allowed to maintain under the current area marking; the air supply regulation boundary corresponds to the adjustable range of air supply temperature, air supply volume, or terminal opening; the setpoint adjustment boundary corresponds to the adjustment direction and range of the thermal environment setpoint; and the lighting and shading linkage boundary corresponds to the linkage limitation between the lighting brightness level and the shading execution level. The minimum hold duration is used to limit the continuous duration of operation after the cooperative control sequence enters. The reverse switching prohibition condition is used to limit the reverse switching when the current sequence hold phase is not completed. The manual intervention hold condition is used to limit the hold state before automatic control resumes after manual intervention is triggered. The control boundary and sequence hold condition are determined based on the area marker and equipment interlocking relationship, and the control boundary and sequence hold condition are used as constraints for calling the control sequence library. When automatic control is first entered, the default control sequence is written into the current cooperative control sequence identifier. During the execution of the current cooperative control sequence, the sequence switching corresponding to the area marker change is only performed after the minimum hold duration is met and the reverse switching prohibition condition is lifted.

[0013] In some embodiments, a protection sequence is selected for the protected area, and the target equipment is controlled in the order of first improving the fresh air capacity and then adjusting the thermal environment; For the adjustment zone, a rapid transition sequence is selected, and the target equipment is controlled in the order of first local adjustment and then overall correction; Select a downgrade sequence for the downgrade zone and control the target equipment in the order of delayed downgrade and graded reduction; Select a conservative sequence for the area to be observed, maintain the current operating mode of the main equipment, and restrict mode switching.

[0014] In some embodiments, the consistency check includes: determining whether the region marker is consistent with the carbon dioxide change trend, temperature and humidity change trend, wireless terminal online change trend and device execution status based on the execution feedback data; generating cross-source conflict results or execution deviation results when inconsistencies exist in multiple consecutive analysis windows; updating the region marker based on the cross-source conflict results or execution deviation results, and re-calling the control sequence library after the region marker changes.

[0015] Secondly, embodiments of the present invention provide a collaborative energy-saving optimization system based on multi-source data analysis and judgment, comprising: The data access module is used to collect appointment scheduling data, access sensing data, wireless access data, indoor environment data, building control data, and equipment operation data. The mapping and profiling module is used to map various types of data to spatial units and their equipment service relationships, and to form a regional operational profile. The analysis and labeling module is used to perform cross-source consistency analysis on the regional operational profile, generate occupancy confidence status, reservation deviation status and indoor air quality conflict status, and determine the regional label of the spatial unit; The boundary generation module is used to generate control boundaries and sequence holding conditions based on the area markers, equipment interlocking relationships, manual intervention holding status, and the current collaborative control sequence. The sequence arrangement module is used to write the default control sequence into the current cooperative control sequence when entering automatic control for the first time, and after the minimum hold time is met and the condition of prohibiting reverse switching is lifted, select a cooperative control sequence from the control sequence library that matches the control boundary and the sequence hold condition; otherwise, maintain the current cooperative control sequence. The hierarchical execution module is used to send the collaborative control sequence to the target device according to the hierarchy of the regional controller and the device controller; The verification rollback module is used to perform consistency verification based on execution feedback data, and to switch to a conservative control sequence and update the region marker when cross-source conflicts persist, execution deviations persist, or manual intervention maintains the status.

[0016] In some embodiments, the system further includes a storage module and a processing module. The storage module stores a space unit mapping table, a device service relationship table, a control sequence library, and a conservative control sequence. The processing module calls the space unit mapping table, the device service relationship table, the control sequence library, and the conservative control sequence.

[0017] Thirdly, embodiments of the present invention provide a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements the collaborative energy-saving optimization method based on multi-source data analysis and judgment as described in the first aspect of the present invention.

[0018] Through the above technical solution, the present invention can achieve at least the following beneficial effects: This invention maps appointment scheduling data, access sensing data, wireless access data, indoor environmental data, building control data, and equipment operation data to spatial units and their equipment service relationships, forming a regional operation profile under a unified time reference. Cross-source consistency analysis is then performed on this regional operation profile to generate occupancy confidence status, appointment deviation status, and indoor air quality conflict status, thereby determining the regional label of the spatial unit. This allows scene recognition to no longer rely on fixed schedules, single access control signals, or low-frequency feedback, but rather on a joint judgment of actual occupancy and environmental response based on multi-source information. This reduces misjudgments of areas under conditions of asynchronous appointment and actual arrival, and lowers the probability of vacant areas being misjudged as areas requiring protection, as well as the probability of temporarily full areas continuing to operate under normal conditions.

[0019] This invention generates control boundaries and sequence holding conditions based on area markings, equipment interlocking relationships, manual intervention holding status, and the current collaborative control sequence. The control sequence library is called only after the minimum holding time is met and the prohibition on reverse switching is lifted. This ensures that the control action is subject to the joint constraints of the fresh air lower limit boundary, air supply adjustment boundary, setpoint adjustment boundary, and lighting and shading linkage boundary before it is issued, as well as the timing constraints of the sequence holding conditions. This reduces frequent back-and-forth switching between adjacent control modes and reduces equipment jitter and prolonged local waiting time caused by false triggering of air supply, fresh air, and cooling.

[0020] This invention divides the area into a protection zone, an adjustment zone, a downgrade zone, and an observation zone, and selects a protection sequence, a rapid transition sequence, a downgrade sequence, and a conservative sequence according to the area marking. This allows high-density spaces, sudden gathering spaces, continuously low-occupancy spaces, and spaces with insufficient evidence to enter different control paths, thereby reducing the ventilation response lag in high-density spaces such as meeting rooms and classrooms, and reducing the ineffective energy consumption caused by continuously maintaining a high protection condition in low-occupancy areas.

[0021] This invention collects execution feedback data and performs consistency verification. When cross-source conflicts persist, execution deviations persist, or manual intervention maintains the status, it switches to a conservative control sequence and updates the region marker. This creates a closed loop for scene identification, control issuance, result verification, and anomaly rollback, thereby reducing the risk of untraceability after manual modification, sensor anomalies, and strategy switching, and reducing the risk of automatic control continuing to spread mismatch under abnormal operating conditions. Attached Figure Description

[0022] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation on the scope of this application.

[0023] Figure 1This is a flowchart of the collaborative energy-saving optimization method based on multi-source data analysis and judgment in the embodiments; Figure 2 This is a framework diagram of the collaborative energy-saving optimization system based on multi-source data analysis and judgment in the embodiment. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0025] All terms used in this application (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0026] Furthermore, the relevant terms and concepts involved in the embodiments are introduced below: A spatial unit refers to the smallest area that serves as a unified object of analysis and control within the same control cycle; equipment service relationship refers to the correspondence between equipment and spatial units in terms of air supply, heat exchange, lighting, or shading; business affiliation relationship refers to the affiliation of a spatial unit with corresponding meeting, teaching, or office business; unified time base refers to the time reference used to align data from different sources along the same time axis; analysis window refers to a continuous time slice divided according to a unified time base and used for status determination; alignment key refers to the index field used to associate records from different sources within the same spatial unit and analysis window, including at least the spatial unit identifier, analysis window identifier, and source identifier. The system establishes a spatial unit mapping table and an equipment service relationship table. The spatial unit mapping table includes at least the spatial unit identifier, floor identifier, area type, business affiliation, adjacent spatial unit identifier, service equipment identifier, controller identifier, validity identifier, and update time; the equipment service relationship table includes at least the equipment identifier, equipment type, service spatial unit identifier, service priority, interlocking object identifier, control path identifier, and status acquisition path identifier. Event-based data is written to the corresponding analysis window according to the event occurrence time; continuously sampled data is aggregated to the corresponding analysis window according to the sampling time; and late cross-system data is backfilled to the corresponding analysis window according to a unified time base, retaining the source identifier and update time. A regional operation profile refers to a set of status descriptions formed by aggregating appointment scheduling data, access sensing data, wireless access data, indoor environment data, building control data, and equipment operation data by spatial unit. Regional operation profiles are generated by spatial unit and continuously updated. Each regional operation profile record includes at least the spatial unit identifier, analysis window identifier, appointment scheduling summary, access sensing summary, wireless access summary, indoor environment summary, building control summary, equipment operation summary, manual intervention status, previous region marker, and current collaborative control sequence identifier.

[0027] Occupancy confidence status, reservation deviation status, and indoor air quality conflict status are used to characterize the actual occupancy confidence level of a space unit, the deviation between the reservation schedule and the actual arrival, and the inconsistency between the occupancy status and changes in ventilation and air parameters, respectively. Area label refers to the control classification formed based on the above status results. The area label judgment rule set refers to the set of rule items used to divide space units into protection zones, adjustment zones, de-configuration zones, and observation zones. Cross-source consistency value refers to the continuous state quantity obtained by synthesizing multiple types of sub-consistencies. Consistency threshold refers to the threshold used to determine whether the cross-source consistency value meets the observation judgment condition. Control boundary refers to the ventilation, thermal environment, and energy consumption adjustment restrictions corresponding to the execution of the coordinated control sequence. Sequence maintenance conditions refer to the timing conditions on which the control sequence is switched, maintained, and exited. Control sequence library refers to the set of coordinated control sequences pre-organized according to area labels, equipment service relationships, and control boundaries. Coordinated control sequence refers to the set of multi-equipment control actions issued to the same space unit or adjacent space units in a predetermined order. Conservative control sequence refers to the restricted control sequence that maintains stable operation when cross-source conflicts persist, execution deviations persist, or manual intervention maintains the status. Target equipment refers to the set of equipment that receives control commands according to the current coordinated control sequence; equipment interlocking relationship refers to the constraint relationship that target equipment needs to jointly satisfy during start-up, shutdown, mode switching, or opening adjustment; execution feedback data refers to the execution status and environmental response data returned by the target equipment and the environmental monitoring terminal; cross-source conflict result refers to the conflict judgment result obtained after consistency verification of multi-source data within the same analysis window; execution deviation result refers to the deviation judgment result formed when the target action and execution feedback data are inconsistent; execution evidence index key refers to the index field used to associate regional operation profile records, control sequence records, and execution evidence records, which includes at least event identifier, spatial unit identifier, analysis window identifier, and version identifier.

[0028] Manual intervention maintenance status refers to the state where manual control has been issued but not yet released, or where it has been released but is still in the recovery and exit phase; default control sequence refers to the control sequence written into the current collaborative control sequence and used as the basis for initial control when entering automatic control for the first time; evidence integrity level refers to the completeness of valid data sources that can be used for cross-source consistency analysis within the current analysis window; current controller lock status refers to the state where the controller is manually locked, policy locked, or interlocked.

[0029] Example 1: This embodiment focuses on the identification and control starting point determination of spatial units under asynchronous multi-source conditions during daytime operation in a mixed-use office and teaching park. This includes: collecting reservation scheduling data, access sensing data, wireless access data, indoor environmental data, building control data, and equipment operation data; and completing spatial unit mapping, unified time reference alignment, regional operation profile generation, cross-source consistency analysis, and regional label output. Based on this, the system then proceeds to the collaborative control sequence selection and consistency verification process. Through this processing chain, the occupancy status, reservation deviation status, and indoor air quality conflict status of spatial units are unified within the same analysis window, ensuring a consistent input basis for subsequent control boundaries and control sequences.

[0030] like Figure 1 As shown, this embodiment proposes a collaborative energy-saving optimization method based on multi-source data analysis and judgment, applied to the daytime operation period of a mixed-use office and teaching park, including: Step S1: Collect reservation scheduling data, access sensing data, wireless access data, indoor environment data, building control data, and equipment operation data, and map these data to spatial units and their equipment service relationships; Step S2: Align the mapped data according to a unified time benchmark, and combine spatial adjacency, equipment service relationship and business affiliation relationship to form a regional operation profile of the spatial unit; Step S3: Perform cross-source consistency analysis on the regional operation profile to generate occupancy confidence status, reservation deviation status, and indoor air quality conflict status, and determine the regional label of the spatial unit; Step S4: Based on the area marker, equipment interlocking relationship, manual intervention hold status, and current collaborative control sequence, generate control boundaries and sequence hold conditions; upon first entering automatic control, write the default control sequence into the current collaborative control sequence; upon first entering automatic control, write the default control sequence into the current collaborative control sequence; generate control boundaries including fresh air lower limit boundary, air supply adjustment boundary, setpoint adjustment boundary, and lighting / shading linkage boundary, as well as sequence hold conditions including minimum hold duration, reverse switching prohibition condition, and manual intervention hold condition; and only after the minimum hold duration is met and the reverse switching prohibition condition is lifted, select a collaborative control sequence matching the control boundaries and sequence hold conditions from the control sequence library and issue it to the target equipment according to the hierarchy of area-level controller and equipment-level controller; if the minimum hold duration is not met or the reverse switching prohibition condition is not lifted, maintain the current collaborative control sequence; Step S5: Collect execution feedback data and verify its consistency with the region label. If any of the following conditions are met, switch to the conservative control sequence and update the region label: cross-source conflict persists; execution deviation persists; manual intervention maintains the status. In this embodiment, the appointment scheduling data includes meeting appointment data, course scheduling data, and temporary scheduling data; Access control data includes access control data and area-to-exit records; Wireless access data includes wireless terminal access status data; Indoor environmental data includes carbon dioxide data, temperature data, and humidity data; Building control data includes damper status data, valve status data, supply and return air status data, and setpoint data; Equipment operation data includes start / stop status data, operation mode data, and fault alarm data for air handling equipment, fresh air equipment, terminal conditioning equipment, lighting equipment, and sunshade equipment; In this embodiment, cross-source consistency analysis is performed, including: Within the same analysis window, the reservation arrival deviation result is obtained based on the reservation scheduling data and access perception data, the occupancy change result is obtained based on the access perception data and wireless access data, the ventilation response result is obtained based on the occupancy change result and indoor environment data, building control data and equipment operation data, and the occupancy confidence state, reservation deviation state and indoor air quality conflict state are determined based on the reservation arrival deviation result, occupancy change result and ventilation response result. In this embodiment, cross-source consistency analysis is performed according to a preset rule set or a state determination model. The state determination model refers to a model that outputs occupancy confidence state, reservation deviation state, and indoor air quality conflict state based on multi-source samples. The training samples for the state determination model are derived from the concurrent operation records of the reservation system, access control system, wireless access system, building control system, and environmental monitoring terminal. The labeled results include at least continuous attendance, reservation attendance deviation, sudden increase in occupancy, continuous low occupancy, ventilation response lag, equipment execution anomaly, and maintenance by manual intervention. The training objective of the state determination model is to ensure that the state results output within the same analysis window are consistent with the actual operation results.

[0031] During the deployment phase, the following data within the same analysis window are used as input: number of appointments, planned start time, planned end time, number of access control entries, number of access control exits, area-to-departure records, number of online wireless terminals, carbon dioxide trend, temperature trend, humidity trend, damper status, valve status, supply and return air status, setpoints, equipment operating mode, and fault alarm status. The outputs are occupancy confidence status, appointment deviation status, and indoor air quality conflict status. Occupancy confidence status includes at least high-confidence occupancy, medium-confidence occupancy, and low-confidence occupancy; appointment deviation status includes at least appointment consistency, late appointment, vacant appointment, and extended appointment; indoor air quality conflict status includes at least no conflict, ventilation lag, and environmental anomaly. If any data source is missing, cross-source consistency analysis retains the missing source's identifier and continues status determination based on the remaining sources, while simultaneously lowering the evidence integrity level of the corresponding analysis window.

[0032] In this embodiment, the reservation arrival deviation result is obtained by comparing the arrival time, departure time, and number of people changes within the same spatial unit and analysis window using reservation scheduling data and access perception data; the occupancy change result is obtained by comparing the magnitude and direction of change between access perception data and wireless access data within the same spatial unit and adjacent analysis windows; the ventilation response result is obtained by comparing the response timing relationship between occupancy change result and indoor environmental data, building control data, and equipment operation data. During operation, historical samples are written to the update set on a daily, weekly, or scene switching cycle, and the status judgment model or preset rule set is updated according to version. The updated version identifier is written to the regional operation profile record and execution evidence record.

[0033] When the state determination model is implemented using a neural network, the neural network refers to a parameterized model that maps multi-source input features to state categories. Training samples are constructed on an analysis window basis, with sample labels directly corresponding to continuous attendance, scheduled attendance deviation, sudden increase in occupancy, continuous low occupancy, delayed ventilation response, equipment execution anomaly, and maintenance by manual intervention. The loss function uses at least one of multi-class cross-entropy, weighted cross-entropy, and focus loss. During deployment, the state category and its confidence level are output for each analysis window. If the confidence level is lower than a preset threshold, it is reviewed according to a preset rule set, and the final state result is output.

[0034] In this embodiment, the area markings include the protection area, adjustment area, downgrade area and observation area, and only one result is retained in the same analysis window; Different spatial types and time periods call up the area marking judgment rule set respectively. The area marking judgment rule set includes at least the attendance ratio judgment item, online change judgment item, carbon dioxide change judgment item, duration judgment item, consistency judgment item, and manual intervention maintenance duration judgment item. Among them, continuous attendance is determined by the joint result of appointment scheduling data, access perception data, and wireless access data; sudden increase in occupancy is determined by the joint result of the change in the number of entries, the change in the number of online terminals, and the carbon dioxide change trend in adjacent analysis windows; continuous low occupancy is determined by the results of low attendance, low online and no indoor air quality conflict in continuous analysis windows; and manual intervention maintenance status is determined by the joint result of manual issuance time, manual release status, and current controller lock status. When a spatial unit is characterized as continuously present within a preset continuous time period and the indoor air quality conflict status is characterized as having a conflict, it is marked as a protected area. When a space unit is characterized as continuously low occupancy and indoor air quality conflict status is characterized as non-conflict within a preset continuous time period, it is marked as a downgraded zone. A spatial unit will be marked as an adjustment zone under any of the following circumstances: the spatial unit is characterized by a sudden increase in occupancy within a preset continuous time period; or the reservation deviation status is characterized by the reservation arrival deviation exceeding a threshold. A spatial unit is marked as an observation area under any of the following conditions: the cross-source consistency of the spatial unit is below the consistency threshold; or manual intervention maintains the status. When a conflict occurs in the area labeling, the final area labeling is determined in the order of observation area, protection area, adjustment area, and de-application area; when the area labeling of the same spatial unit changes within a continuous analysis window, the sequence preservation condition is verified first, and then the area labeling is updated. In this embodiment, the meeting space, teaching space, and open office space respectively call the corresponding area label determination rule set, and only one result of the area label is retained in the same analysis window; when the area label conflicts, the final area label is determined in the order of the observation area, the protection area, the adjustment area, and the downgrade area; when the area label changes in the same spatial unit in the continuous analysis window, the sequence preservation condition is checked first, and then the area label update is performed.

[0035] In an optional implementation of Example 1, if the cross-source consistency of a spatial unit is lower than the consistency threshold, it is determined one spatial unit at a time, with an analysis window as the rolling calculation period. For each spatial unit, four types of sub-consistency are formed within the same analysis window: appointment-arrival consistency, arrival-online consistency, online-environment consistency, and environment-equipment response consistency, which are then combined into a cross-source consistency value. , in, The spatial unit is represented in the first place. Cross-source consistency values ​​within each analysis window; Indicates spatial unit identifier; Indicates the analysis window number; Indicates the sub-consistency category number. Consistency between appointment and attendance. Corresponding on-site and online consistency Corresponding to online-environment consistency, Corresponding environment—consistency of device response; Indicates the first Effective weights for class consistency; Indicates the first The evaluation value for sub-consistency ranges from 0 to 1, with smaller values ​​indicating stronger conflicts. Linear weighted aggregation is only performed when there are at least three valid sub-consistency categories. When the evaluation value of any sub-consistency is lower than the strong conflict threshold for the corresponding sub-consistency, and the corresponding source availability coefficient is not lower than the source reliability threshold for the corresponding sub-consistency, this sub-consistency category is directly recorded as a conflict category, and its conflict attribute is not removed due to higher evaluation values ​​of other sub-consistencies. When the number of conflict categories in the current analysis window reaches two, the conflict determination path is prioritized, and normal consistency determination is no longer made solely based on the linear aggregation result. The strong conflict threshold and source reliability threshold are calibrated offline according to sub-consistency category, business type, and time period, and remain unchanged within the same continuous observation window.

[0036] The calculation of various sub-consistencies adopts a unified standard for deviation magnitude, direction matching, and duration. Reservation-Attendance Consistency compares the corrected number of reservations with the actual number of attendees within the analysis window, while also considering the allowable attendance offset before and after the reservation start time. When the proportion of the number of attendees deviates from the corresponding service type's attendance deviation reference value, and the attendance lag falls outside the lag interval for multiple consecutive analysis windows, this type of consistency is considered to have decreased. Attendance-Online Consistency compares regional arrival and departure records with the deduplicated number of wireless online terminals, excluding fixed network terminals, roaming duplicate terminals, and short-term dwell terminals. When the direction of attendance change is opposite to the direction of online change, or when the proportion of deviation after adjusting the number of attendees consistently exceeds the online deviation reference value, this type of consistency is considered to have decreased. Online-Environment Consistency correlates changes in online terminals with the response trends of carbon dioxide, temperature, and humidity, using the environmental change amount after delay compensation as the comparison object. When the number of online attendees continues to increase while carbon dioxide and thermal / humidity disturbances do not show corresponding changes within the specified response lag, or when the number of online attendees continues to decrease while environmental indicators continue to deviate from the reference band, this type of consistency is considered to have decreased. Environmental-equipment response consistency is based on the matching relationship between environmental deviation and the direction of equipment execution feedback. When carbon dioxide or thermal environmental indicators exceed the control dead zone, if the fresh air, supply and return air, valves, or terminal control equipment fail to change in the target direction within the response time lag corresponding to the equipment service relationship, or if the equipment execution status has changed but the environmental improvement is continuously lower than the minimum response amplitude, this type of consistency is judged to have decreased. The reference values ​​for the four types of sub-consistencies are all calibrated according to the business type, time period, and historical normal operation samples of the same spatial unit. Among them, the on-site deviation reference value, online deviation reference value, environmental response band, and minimum equipment response amplitude are all taken from the quantile or mean band of the recent open window samples. The values ​​are adjusted with the historical fluctuation level. A larger tolerance band is used for spatial units with high fluctuations, and a tighter tolerance band is used for spatial units with low fluctuations. When the historical normal operation sample size of the same spatial unit under the corresponding business type and time period is lower than the preset minimum sample size, the on-site deviation reference value, online deviation reference value, environmental response band, and minimum equipment response amplitude will not be recalculated, and the previous valid parameter version will be used first; if the previous valid parameter version does not exist, the park-level conservative reference value of the same business type and time period will be called. New samples formed during the period of insufficient samples are only used to supplement the sample size and do not directly trigger the update of reference statistics.

[0037] To ensure the feasibility of consistency determination under source-deficient conditions, the weights of the four types of sub-consistencies are handled by degrading according to source availability: , in, Indicates the first Subclass consistency in the first The source availability coefficient within each analysis window ranges from 0 to 1. Indicates the first The basic weights for class consistency; This represents the category count index in the normalization calculation. When a data source is completely missing, timestamp drift exceeds the limit, or values ​​are stuck, the corresponding source availability coefficient drops to 0. When a data source only has short-term delays, local packet loss, or increased noise, the corresponding source availability coefficient drops to a reduced value between 0 and 1, and the remaining effective weights are redistributed among the remaining sub-consistencies. The basic weights are reorganized according to the business type using preset weights. In the teaching scheduling space, the basic weights for appointment-attendance consistency and environment-equipment response consistency are taken as higher values. In the flexible office space, the basic weights for attendance-online consistency and online-environment consistency are taken as higher values. If the sum of the effective weights before normalization is lower than the preset weight lower limit, it indicates that there is insufficient effective evidence in the current analysis window that can be used for cross-source verification. In this case, continuous cross-source consistency values ​​are not output, and instead, a conservative threshold group is called to perform the observation judgment.

[0038] If, within an analysis window that continuously reaches the preset observation window length, two or more data sources corresponding to sub-consistency are unavailable, or the effective weights before normalization are... If the value remains consistently below the preset weight lower limit, normal consistency judgment will no longer be maintained, and the conservative threshold group will be directly invoked. If the weight normalization calculation is not performed, the current analysis window is directly deemed unable to form an effective cross-source consistency judgment, and the spatial unit is marked as an observation area according to a conservative strategy.

[0039] The consistency threshold is not a fixed value, but rather a threshold group is called according to business type, time period, and historical fluctuation level, and amplitude limit correction is performed in combination with the degree of sensor anomaly: , in, The spatial unit is represented in the first place. Consistency thresholds used within each analysis window; Indicates by business type and time period A jointly selected reference threshold; This represents the historical volatility correction factor; Indicates the historical fluctuation level of a spatial unit; Indicates the anomaly correction factor; Indicates the degree of sensor abnormality; This indicates the lower limit of the consistency threshold; This indicates the upper limit of the consistency threshold; This indicates amplitude limiting processing. Historical fluctuation levels are calibrated by the cross-source consistency dispersion of the same spatial unit under the same business type and similar time period. The greater the fluctuation, the lower the reference threshold is adjusted to avoid frequent entry into the observation zone due to short-term disturbances. Sensor anomaly severity is comprehensively calibrated by the proportion of missing reports, lag duration, jump variables, and number of verification failures. The more severe the anomaly, the higher the reference threshold is adjusted to trigger the observation judgment earlier. For spatial units in a manually intervened state, the threshold comparison process is skipped, and they are directly marked as the observation zone. Within the preset release holding window after the manual intervention is lifted, the conservative threshold group is called to control the risk of misjudgment during the recovery phase. The preset release holding window is calibrated according to the spatial unit's business type and historical switching frequency. Spatial units with higher switching frequencies take a longer release holding window, while spatial units with lower switching frequencies take a shorter release holding window.

[0040] Reference threshold groups, basic weighted reassemblies, environmental response bands, and minimum equipment response amplitudes are loaded with the same parameter version and remain unchanged within the same continuous observation window. When in a manual intervention hold state, a preset release hold window, or a recovery exit phase, parameter version switching is not performed; the previous valid parameter version continues to be used. If the relationship between the lower and upper thresholds of the current parameter version is abnormal, the new version of the parameters is not called; the previous valid parameter version is directly used, and the conservative threshold group is invoked.

[0041] When the cross-source consistency value is lower than the consistency threshold and the persistence condition is met, it is considered that the cross-source consistency is lower than the consistency threshold. , in, Indicates the spatial unit up to the th The cumulative number of low consistency occurrences for each analysis window; Indicates the sequence number of the analysis window used for forward backtracking; Indicates the length of the continuous observation window; This represents an indicator function, which takes the value 1 when the condition is true and 0 when the condition is false. The spatial unit is represented in the first place. A cross-source consistency output validity flag is set within each analysis window. A value of 1 indicates that the current analysis window outputs a valid cross-source consistency value, while a value of 0 indicates that the current analysis window does not output a cross-source consistency value due to a missing source rollback. When the current analysis window is in a state of low consistency, the cumulative count is directly calculated based on the low consistency window value, and the cross-source consistency value of the previous analysis window is not used as a replacement value; when When the continuous observation window length is reached, and at least two of the four types of sub-consistencies are in conflict within the same continuous observation window, the cross-source consistency of the spatial unit is determined to be below the consistency threshold, and the spatial unit is marked as a region to be observed. When the environment-device response consistency shows that the device execution direction is opposite to the environment deviation direction for at least two consecutive analysis windows, or the device execution state has changed but the improvement in the environment is continuously lower than the minimum response amplitude of the device, the spatial unit can be directly determined to enter the region to be observed without waiting for the cumulative number of times to reach the continuous observation window length. The above processing method ensures that the entry conditions for the region to be observed simultaneously have the basis for cross-source conflict, the basis for the duration, and the conservative basis under abnormal operating conditions, thus maintaining consistency with the region marking logic. Implementation parameters are loaded according to the same parameter version. The analysis window can be 5 minutes, with a value range of 1 to 15 minutes; the continuous observation window can have 3 analysis windows; the reduction value of the source availability coefficient can be 0.5 to 0.8; the preset weight lower limit can be 0.35 to 0.55; the reference threshold, environmental response band, and minimum response amplitude of the equipment can be adjusted based on the quantile of the normal operation sample over the past 30 days and remain unchanged within the same continuous observation window; when there are insufficient historical samples, the previous effective parameter version is used; when the previous effective parameter version is missing, the park-level conservative reference value is called.

[0042] In this embodiment, a protection sequence is selected for the protection zone, and the target equipment is controlled in the order of first improving the fresh air capacity and then adjusting the thermal environment; For the adjustment zone, a rapid transition sequence is selected, and the target equipment is controlled in the order of first local adjustment and then overall correction; Select a downgrade sequence for the downgrade zone and control the target equipment in the order of delayed downgrade and graded reduction; For the area to be observed, a conservative sequence is selected to maintain the current operating mode of the main equipment and restrict mode switching; In this embodiment, each collaborative control sequence in the control sequence library corresponds one-to-one or many-to-one with a region marker. Each collaborative control sequence includes at least a sequence identifier, applicable region marker, target device list, action sequence, phase duration, entry condition, exit condition, rollback condition, interlock restrictions, and version identifier. The safeguard sequence includes a fresh air capacity enhancement phase, a thermal environment correction phase, and a stable maintenance phase; the rapid transition sequence includes a local capacity enhancement phase, a cross-source verification phase, and an overall correction phase; the capacity reduction sequence includes a delayed capacity reduction phase, a tiered phasing phase, and a low-load maintenance phase; and the conservative sequence includes a current mode maintenance phase, a conflict verification phase, and a manual release waiting phase.

[0043] In this embodiment, hierarchical control is executed with regional controllers taking precedence, followed by device-level controllers. After receiving the regional marker, control boundary, and collaborative control sequence identifier corresponding to the spatial unit, the regional controller generates a target device list and phase sequence. The device-level controller executes start / stop, mode switching, setpoint adjustment, valve opening adjustment, damper opening adjustment, lighting brightness adjustment, and shading according to the target device list and phase sequence. When adjacent spatial units share the same service equipment, service priority is first determined in the order of protection zone, adjustment zone, downgrade zone, and observation zone, and then device-level control commands are generated by merging them according to control boundaries.

[0044] Device-level control command records refer to the merged output records of shared service devices within the same analysis window, including at least the device identifier, spatial unit identifier set, priority result, target action, limiting result, and effective time. When shared service devices receive conflicting target actions, they first determine the target action according to the area tag priority, and then limit the opening degree, setpoint, and start / stop result according to the control boundary. If conflicts still exist after limiting, the operating mode of the main device corresponding to the current cooperative control sequence is maintained and a conflict identifier is written.

[0045] In this embodiment, consistency verification includes: judging whether the region marker is consistent with the carbon dioxide change trend, temperature and humidity change trend, wireless terminal online change trend and device execution status based on the execution feedback data; when there is inconsistency in multiple consecutive analysis windows, generating cross-source conflict results or execution deviation results; updating the region marker based on the cross-source conflict results or execution deviation results, and re-calling the control sequence library after the region marker changes; In this embodiment, consistency verification focuses on the correspondence between area markings, execution results, and feedback changes. When the area is marked as a protection zone, the focus is on verifying the consistency between the carbon dioxide change trend, the air supply or fresh air execution status, and the equipment operating mode; when the area is marked as an adjustment zone, the focus is on verifying the consistency between changes in the number of online terminals, changes in access, and localized expansion actions; when the area is marked as a reduction zone, the focus is on verifying the consistency between low occupancy status, stable environmental status, and phase-out actions; when the area is marked as an observation zone, the focus is on verifying the consistency between manual intervention maintenance status, equipment lock status, and current mode maintenance status. Persistent cross-source conflicts are determined by the unresolved state conflicts within the continuous analysis window; persistent execution deviations are determined by the inconsistency between the target action and equipment feedback within the continuous analysis window.

[0046] In this embodiment, execution evidence records refer to data records used to document the entire process of assessment, issuance, execution, verification, and rollback. Each execution evidence record includes at least the event identifier, spatial unit identifier, analysis window identifier, region marker, collaborative control sequence identifier, target device list, issuance time, feedback time, cross-source conflict result, execution deviation result, manual intervention maintenance status, conservative control sequence identifier, version identifier, and update time. Execution evidence records are continuously written according to the analysis window, used to trace the basis for region marker formation, control boundary source, collaborative control sequence switching path, and conservative control sequence triggering reasons. After the cross-source conflict result or execution deviation result is resolved, the system first completes the maintenance phase of the current conservative control sequence before resuming calls to the control sequence library.

[0047] The execution evidence storage index key refers to the index field used to associate the region's operational profile record, control sequence record, and execution evidence storage record. It includes at least the event identifier, spatial unit identifier, analysis window identifier, and version identifier. When control command issuance fails or communication is interrupted, the most recently valid issuance record is cached according to the analysis window, and the current collaborative control sequence is maintained. After communication is restored, the cached record is rewritten first, and then consistency verification and region tag updates continue.

[0048] Example 2: Based on Example 1, this example details the control boundary convergence and sequence switching after the area markers are determined. This includes generating fresh air lower limit boundaries, supply air adjustment boundaries, setpoint adjustment boundaries, lighting and shading linkage boundaries, and corresponding sequence holding conditions based on area markers, equipment interlocking relationships, manual intervention hold states, and the current collaborative control sequence. Then, it continuously determines the conditions prohibiting reverse switching and constrains the entry, maintenance, exit, and recovery of the control sequence accordingly. Through this processing chain, the control actions are seamlessly integrated with equipment protection constraints, scene recovery states, and manual intervention hold states, ensuring stable switching criteria for the control sequence within adjacent analysis windows.

[0049] In this embodiment, the control boundaries include a fresh air lower limit boundary, an air supply regulation boundary, a setpoint adjustment boundary, and a lighting and shading linkage boundary. The sequence maintenance conditions include a minimum maintenance duration, a prohibition on reverse switching conditions, and a manual intervention maintenance condition. Among them, the fresh air lower limit boundary corresponds to the lowest fresh air capacity level that the spatial unit is allowed to maintain under the current area marking; the air supply regulation boundary corresponds to the adjustable range of air supply temperature, air supply volume, or terminal opening; the setpoint adjustment boundary corresponds to the adjustment direction and adjustment range of the thermal environment setpoint; and the lighting and shading linkage boundary corresponds to the linkage restriction between the lighting brightness level and the shading execution level. The minimum hold duration is used to limit the continuous duration of operation after the cooperative control sequence is entered. The reverse switching prohibition condition is used to limit the reverse switching when the current sequence hold phase is not completed. The manual intervention hold condition is used to limit the hold state before automatic control resumes after manual intervention is triggered. The control boundary and sequence hold condition are determined based on the area marker and equipment interlocking relationship, and the control boundary and sequence hold condition are used as constraints for calling the control sequence library. When entering automatic control for the first time, the default control sequence is written into the current cooperative control sequence identifier. During the execution of the current cooperative control sequence, the sequence switching corresponding to the area marker change is only performed after the minimum hold duration is met and the reverse switching prohibition condition is lifted. In this embodiment, the fresh air lower limit boundary corresponds to the lowest fresh air capacity level that a space unit is allowed to maintain under the current area marking; the supply air adjustment boundary corresponds to the adjustable range of supply air temperature, supply air volume, or terminal opening; the setpoint adjustment boundary corresponds to the adjustment direction and adjustment range of the thermal environment setpoint; and the lighting and shading linkage boundary corresponds to the linkage restriction between the lighting brightness level and the shading execution level. The minimum hold duration is used to limit the continuous duration of operation after the collaborative control sequence enters; the prohibition of reverse switching condition is used to limit reverse switching when the current sequence hold phase is not completed; and the manual intervention hold condition is used to limit the hold state before automatic control recovery after manual intervention is triggered. The record content corresponding to the control boundary includes at least the space unit identifier, analysis window identifier, area marking, fresh air lower limit boundary, supply air adjustment boundary, setpoint adjustment boundary, lighting and shading linkage boundary, interlock object identifier, applicable equipment identifier, and version identifier. The record content corresponding to the sequence hold condition includes at least the space unit identifier, current collaborative control sequence identifier, minimum hold duration, prohibition of reverse switching condition, manual intervention hold condition, entry time, and exit condition. When entering automatic control for the first time, the default control sequence is written into the current cooperative control sequence identifier; during the execution of the current cooperative control sequence, the sequence switching corresponding to the change of the area marker is only performed after the minimum hold time is met and the condition of prohibiting reverse switching is lifted.

[0050] In one optional implementation of Example 2, the reverse switching prohibition condition is used to constrain spatial units from switching back and forth along opposite adjustment directions within adjacent analysis windows. The main adjustment direction corresponding to the current cooperative control sequence is compared with the main adjustment direction corresponding to the candidate control sequence. When the core control quantity of the same spatial unit changes from increasing to decreasing, from decreasing to increasing, or from the currently maintained direction to an adjustment direction opposite to the current control boundary, it is identified as a reverse switching request. Core control quantities include fresh air capacity, supply air adjustment capacity, temperature setpoint, lighting output, and shading opening. The reverse switching prohibition condition is not triggered solely by changes in area markers, but rather uses equipment protection constraints, scene recovery criteria, and manual intervention priority together as a release gating. If any gating is not released, the current cooperative control sequence is maintained or switched to a conservative control sequence; candidate reverse sequences are not allowed to be executed.

[0051] For equipment protection constraints, the minimum holding time, action completion confirmation, and interlock release conditions of each key actuator participating in this round of adjustment within the space unit are checked to form the equipment protection release status: , in, The spatial unit is represented in the first place. The device protection release status is displayed in the analysis window, with a value of 0 or 1. Indicates spatial unit identifier; Indicates the analysis window number; Indicates the serial number of the critical execution device; This indicates the total number of key actuators within the space unit that participated in this round of regulation; This represents an indicator function, which takes the value 1 when the condition is true and 0 when the condition is false. Indicates the first The duration of continuous hold of each key actuator since the most recent directional adjustment was completed; Indicates the first Minimum lock-hold duration for each key execution device; Indicates the first Confirmation status of the completion of actions of key execution devices; Indicates the first The interlock release status corresponding to each key execution device.

[0052] The minimum lock-in duration is calibrated separately for each equipment type during the commissioning phase. For air valves and dampers, the reference caliber is the length of the preset stable analysis window after they reach their designated position. For fans, the reference caliber is the length of the preset stable analysis window after the fan speed or air volume enters the stable zone. For end effectors, the reference caliber is the length of the opening position and the rate of displacement change falling back into the stable zone. When a space unit is associated with direct expansion equipment or other equipment with compressors, the reference caliber is the combined start-stop protection duration and restart waiting time. Action completion confirmation is determined based on execution feedback data. When the deviation between the equipment feedback position and the target position falls within the preset confirmation zone and continuously maintains the preset action confirmation analysis window length, the action completion confirmation status is set to 1. The interlock release status is determined based on equipment service relationships and interlock relationships. When there is no mutual exclusion, fault interlock, or safety interlock between supply and return air, fresh air, valves, and end equipment, the interlock release status is set to 1. If any critical actuator fails to reach the minimum lock-in duration, the action is not completed, or the interlock is not released, the equipment protection release status is set to 0, and reverse switching is prohibited.

[0053] For scene recovery criteria, a single change in area markers is not used as the basis for reverse switching. Instead, within the recovery observation window, three conditions are simultaneously verified: occupancy change decline, indoor air quality conflict resolution, and stable equipment feedback, to form a scene recovery release state. , in, The spatial unit is represented in the first place. The scene recovery release state of each analysis window, with a value of 0 or 1; This indicates the recovery of the occupancy change residual within the observation window; This indicates the upper limit of the residual corresponding to the fallback criterion. This indicates the status of indoor air quality conflict; a value of 0 indicates that the conflict has been resolved, and a value of 1 indicates that the conflict still exists. This indicates the amount of equipment feedback fluctuation within the recovery observation window; This indicates the upper limit of fluctuation corresponding to the device feedback stability criterion. The occupancy change residual is the non-negative normalized result of the difference between the changes in the number of attendees, the changes in wireless online terminals, and the recently revised number of appointments within the recovery observation window. The normalized denominator is the larger value between the spatial unit service capacity and the recently revised number of appointments. The device feedback fluctuation is the non-negative normalized result of the opening deviation, speed deviation, start / stop jitter count, and setpoint deviation in the execution feedback data relative to the device's allowable range. Both the occupancy change residual and the device feedback fluctuation are limited to the range of 0 to 1. If any input data is missing, the timestamp is misaligned, or the feedback value is stuck, the corresponding quantity is treated as exceeding its upper limit, and the scene recovery release state is maintained at 0. The occupancy fallback criterion calls the corresponding threshold group according to business type, time period, and historical fluctuation level. A wider fallback band is used for teaching spaces with frequent course switching, and a tighter fallback band is used for stable office spaces. The determination of indoor air quality conflict resolution is based on the return of carbon dioxide to the permissible air quality range, and the return of temperature and humidity to the permissible thermal environment range, while maintaining the preset recovery observation and analysis window length continuously. If the equipment still exhibits back-and-forth shaking, execution lag, or feedback jumps, it is not considered stable.

[0054] The scene recovery release state is set to 1 only when the occupancy change falls back, the indoor air quality conflict is resolved, and the equipment feedback is stable. If any of these conditions are not met, reverse switching will continue to be prohibited.

[0055] For manual intervention priority, automatic control does not participate in reverse handover release during the manual intervention period; after manual release, it does not directly return to the normal control sequence, but enters the recovery exit phase. Before the recovery exit phase is completed, the manual release gating is still regarded as not released. , in, The spatial unit is represented in the first place. The manual release gating status of each analysis window, with a value of 0 or 1; This indicates the status of manual intervention; a value of 1 indicates that manual intervention is in progress, and a value of 0 indicates that manual intervention has been lifted. This indicates the recovery exit completion status; a value of 1 indicates recovery exit is complete, and a value of 0 indicates recovery exit is incomplete. Both the manual intervention status and the recovery exit completion status are refreshed synchronously with the analysis window, and are only updated when two consecutive analysis windows receive consistent status values. If the status value fluctuates, is lost, or synchronization is incomplete, the previous valid status remains unchanged, and reverse switching continues to be prohibited. During the recovery exit phase, the current control sequence version is frozen, and switching to a new normal control sequence version is not performed during the exit process. After the recovery exit is completed, the normal sequence call is restored according to the valid parameter version loaded in the current analysis window.

[0056] The recovery and exit phase is executed step-by-step in the following order: maintaining the current main equipment operating mode, releasing local limits, recovering control boundaries, and restoring the normal sequence. The recovery and exit completion status remains at 0 until any level of exit is completed. The update frequency of each exit step is synchronized with the analysis window, and each exit step is subject to equipment protection constraints. If, during the recovery and exit phase, there is a recurrence of occupancy rebound, air quality conflict, equipment feedback instability, or new manual intervention, subsequent exit steps are suspended, and the current sequence is maintained or a conservative control sequence is initiated.

[0057] Based on the above three types of gating, the conditions for prohibiting reverse switching are determined as follows: , in, The spatial unit is represented in the first place. The state of the analysis window that is prohibited from reverse switching, with a value of 0 or 1; when When, it indicates that the execution of the candidate reverse control sequence is prohibited; when When this is the case, it indicates that the candidate reverse control sequence is allowed to be executed within the current control boundary.

[0058] In practice, the reverse switching prohibition status is updated on a scrolling basis according to the analysis window. If equipment feedback is missing, sensor data is abnormal, interlock status is unclear, or manual status changes are not synchronized, the reverse switching prohibition status is set to 1 according to a conservative strategy, and the current main equipment operating mode remains unchanged; reverse switching is reopened only after the anomaly is resolved and the system is continuously verified through the recovery observation window. Thus, the reverse switching prohibition condition is linked to both equipment lifespan protection and interlock safety constraints, as well as scene restoration realism and manual intervention priority, thereby suppressing frequent back-and-forth switching between adjacent modes. When execution feedback data is missing, interlock status is not refreshed, or manual status is not synchronized, the reverse switching prohibition status remains effective; the corresponding spatial unit maintains the current collaborative control sequence, and the switching determination is restored only after two consecutive analysis windows receive consistent statuses. If occupancy rebound, indoor air quality conflict reappears, or equipment feedback becomes unstable again during the recovery exit phase, subsequent exits are suspended and the current collaborative control sequence is maintained.

[0059] Example 3: Based on Example 2, such as Figure 2As shown, this embodiment is applicable to the systematic deployment and closed-loop execution of collaborative energy-saving optimization methods. It includes: a data access module, a mapping and profiling module, a judgment and marking module, a boundary generation module, a sequence arrangement module, a hierarchical execution module, and a verification and rollback module sequentially completing data access, state judgment, control generation, hierarchical distribution, feedback verification, and conservative rollback. A storage module and a processing module continuously maintain the regional operation profile records, control sequence records, and execution evidence records. This ensures that the data flow, judgment flow, and execution flow in the method form a corresponding relationship within the system, and that execution results and trigger sources can be continuously recorded according to the analysis window.

[0060] Specifically, this embodiment provides a collaborative energy-saving optimization system based on multi-source data analysis and judgment, applied to the daytime operation of a mixed-use office and teaching park, including: The data access module is used to collect appointment scheduling data, access sensing data, wireless access data, indoor environment data, building control data, and equipment operation data. The mapping and profiling module is used to map various types of data to spatial units and their equipment service relationships, and to form a regional operational profile. The analysis and labeling module is used to perform cross-source consistency analysis on the regional operation profile, generate occupancy confidence status, reservation deviation status and indoor air quality conflict status, and determine the regional label of the spatial unit; The boundary generation module is used to generate control boundaries and sequence holding conditions based on area markings, equipment interlocking relationships, manual intervention holding status, and the current collaborative control sequence or default control sequence. The sequence orchestration module is used to select cooperative control sequences from the control sequence library that match the control boundaries and sequence preservation conditions; The hierarchical execution module is used to issue collaborative control sequences to the target device according to the hierarchy of the regional controller and the device controller; The verification rollback module is used to perform consistency verification based on execution feedback data, and to switch to a conservative control sequence and update the region marker when cross-source conflicts persist, execution deviations persist, or manual intervention maintains the status. In this embodiment, the data access module, mapping and profiling module, analysis and marking module, boundary generation module, sequence arrangement module, hierarchical execution module, and verification and rollback module are deployed in at least one of the following: a park management server, a building control server, or an edge control gateway. The storage module stores a space unit mapping table, a device service relationship table, a control sequence library, conservative control sequences, regional operation profile records, and execution evidence records. The processing module calls the data records and sequence records in the storage module and performs data alignment, status determination, regional marking, boundary generation, sequence arrangement, hierarchical distribution, consistency verification, and conservative rollback in a cyclical manner according to the analysis window. When the computer program stored in the computer-readable storage medium is executed by the processor, the processor completes the collaborative energy-saving optimization control of the daytime operation phase of the hybrid office and teaching park according to the above process.

[0061] This embodiment also provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, it implements the collaborative energy-saving optimization method based on multi-source data analysis and judgment proposed in the above embodiment. The storage medium is a volatile or non-volatile storage device.

[0062] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0063] Furthermore, those skilled in the art will understand that although some embodiments herein include certain features included in other embodiments but not others, combinations of features from different embodiments are meant to be within the scope of this application and form different embodiments. For example, all the embodiments above can be used in any combination. The information disclosed in this background section is intended only to enhance the understanding of the general background of this application and should not be construed as an admission or in any way implying that such information constitutes prior art known to those skilled in the art.

Claims

1. A collaborative energy-saving optimization method based on multi-source data analysis and judgment, applied to the daytime operation phase of a mixed-use office and teaching park, characterized in that, include: Step S1: Collect reservation scheduling data, access sensing data, wireless access data, indoor environment data, building control data, and equipment operation data, and map these data to spatial units and their equipment service relationships; Step S2: Align the mapped data according to a unified time benchmark, and combine spatial adjacency, equipment service relationship and business affiliation relationship to form a regional operation profile of the spatial unit; Step S3: Perform cross-source consistency analysis on the regional operation profile to generate occupancy confidence status, reservation deviation status and indoor air quality conflict status, and determine the regional label of the spatial unit; Step S4: Generate control boundaries and sequence holding conditions based on the area markers, equipment interlocking relationships, manual intervention holding status, and the current collaborative control sequence; upon first entering automatic control, write the default control sequence into the current collaborative control sequence; only after the minimum holding time is met and the prohibition of reverse switching is lifted, select a collaborative control sequence matching the control boundaries and sequence holding conditions from the control sequence library, and issue it to the target device according to the hierarchy of area-level controllers and equipment-level controllers; otherwise, maintain the current collaborative control sequence. Step S5: Collect execution feedback data and verify its consistency with the region marker. If any of the following conditions are met, switch to the conservative control sequence and update the region marker: cross-source conflict persists; execution deviation persists; manual intervention maintains the status.

2. The collaborative energy-saving optimization method based on multi-source data analysis and judgment according to claim 1, characterized in that, The appointment scheduling data includes meeting appointment data, course scheduling data, and temporary scheduling data; The access control data includes access control data and area-to-departure records; The wireless access data includes wireless terminal access status data; The indoor environmental data includes carbon dioxide data, temperature data, and humidity data; The building control data includes damper status data, valve status data, supply and return air status data, and setpoint data; The equipment operation data includes start / stop status data, operation mode data, and fault alarm data for air handling equipment, fresh air equipment, terminal control equipment, lighting equipment, and sunshade equipment.

3. The collaborative energy-saving optimization method based on multi-source data analysis and judgment according to claim 1, characterized in that, The cross-source consistency analysis includes: Within the same analysis window, the appointment arrival deviation result is obtained based on the appointment scheduling data and access perception data; the occupancy change result is obtained based on the access perception data and wireless access data; the ventilation response result is obtained based on the occupancy change result, indoor environment data, building control data, and equipment operation data; and the occupancy confidence state, appointment deviation state, and indoor air quality conflict state are determined based on the appointment arrival deviation result, occupancy change result, and ventilation response result.

4. The collaborative energy-saving optimization method based on multi-source data analysis and judgment according to claim 1, characterized in that, The area markings include a protection area, an adjustment area, a downgrade area, and an observation area, and only one result is retained within the same analysis window; Different spatial types and different time periods call the area marking judgment rule set respectively. The area marking judgment rule set includes at least the attendance ratio judgment item, online change judgment item, carbon dioxide change judgment item, duration judgment item, consistency judgment item, and manual intervention maintenance duration judgment item. Among them, continuous attendance is determined by the joint result of appointment scheduling data, access perception data, and wireless access data; sudden increase in occupancy is determined by the joint result of the change in the number of entries, the change in the number of online terminals, and the carbon dioxide change trend in adjacent analysis windows; continuous low occupancy is determined by the results of low attendance, low online and no indoor air quality conflict in continuous analysis windows; and manual intervention maintenance status is determined by the joint result of manual issuance time, manual release status, and current controller lock status. When a spatial unit is characterized as continuously present within a preset continuous time period and the indoor air quality conflict status is characterized as having a conflict, it is marked as a protected area. When a space unit is characterized as continuously low occupancy and indoor air quality conflict status is characterized as non-conflict within a preset continuous time period, it is marked as a downgraded zone. A spatial unit will be marked as an adjustment zone under any of the following circumstances: the spatial unit is characterized by a sudden increase in occupancy within a preset continuous time period; or the reservation deviation status is characterized by the reservation arrival deviation exceeding a threshold. A spatial unit is marked as an observation area under any of the following conditions: the cross-source consistency of the spatial unit is below the consistency threshold; or manual intervention maintains the status. When a conflict occurs in the region label, the final region label is determined in the order of the observation region, the protection region, the adjustment region, and the downgrade region. When the region label changes within the same spatial unit in a continuous analysis window, the sequence preservation condition is verified first, and then the region label is updated.

5. The collaborative energy-saving optimization method based on multi-source data analysis and judgment according to claim 1, characterized in that, The control boundaries include a fresh air lower limit boundary, an air supply regulation boundary, a setpoint adjustment boundary, and a lighting and shading linkage boundary. The sequence holding conditions include a minimum holding time, a prohibition on reverse switching conditions, and a manual intervention holding condition. Among them, the fresh air lower limit boundary corresponds to the lowest fresh air capacity level that the space unit is allowed to maintain under the current area marking; the air supply regulation boundary corresponds to the adjustable range of air supply temperature, air supply volume, or terminal opening; the setpoint adjustment boundary corresponds to the adjustment direction and adjustment range of the thermal environment setpoint; and the lighting and shading linkage boundary corresponds to the linkage restriction between the lighting brightness level and the shading execution level. The minimum hold duration is used to limit the continuous duration of operation after the cooperative control sequence enters. The reverse switching prohibition condition is used to limit the reverse switching when the current sequence hold phase is not completed. The manual intervention hold condition is used to limit the hold state before automatic control resumes after manual intervention is triggered. The control boundary and sequence hold condition are determined based on the area marker and equipment interlocking relationship, and the control boundary and sequence hold condition are used as constraints for calling the control sequence library. When automatic control is first entered, the default control sequence is written into the current cooperative control sequence identifier. During the execution of the current cooperative control sequence, the sequence switching corresponding to the area marker change is only performed after the minimum hold duration is met and the reverse switching prohibition condition is lifted.

6. The collaborative energy-saving optimization method based on multi-source data analysis and judgment according to claim 1, characterized in that, Select a protection sequence for the protected area, and control the target equipment in the order of first improving the fresh air capacity and then adjusting the thermal environment; For the adjustment zone, a rapid transition sequence is selected, and the target equipment is controlled in the order of first local adjustment and then overall correction; Select a downgrade sequence for the downgrade zone and control the target equipment in the order of delayed downgrade and graded reduction; Select a conservative sequence for the area to be observed, maintain the current operating mode of the main equipment, and restrict mode switching.

7. The collaborative energy-saving optimization method based on multi-source data analysis and judgment according to claim 1, characterized in that, The consistency check includes: Based on the execution feedback data, determine whether the region marker is consistent with the carbon dioxide change trend, temperature and humidity change trend, wireless terminal online change trend, and device execution status; when there is inconsistency in multiple consecutive analysis windows, generate cross-source conflict results or execution deviation results; update the region marker based on the cross-source conflict results or execution deviation results, and re-call the control sequence library after the region marker changes.

8. A collaborative energy-saving optimization system based on multi-source data analysis and judgment, based on the collaborative energy-saving optimization method based on multi-source data analysis and judgment as described in any one of claims 1 to 7, characterized in that, include: The data access module is used to collect appointment scheduling data, access sensing data, wireless access data, indoor environment data, building control data, and equipment operation data. The mapping and profiling module is used to map various types of data to spatial units and their equipment service relationships, and to form a regional operational profile. The analysis and labeling module is used to perform cross-source consistency analysis on the regional operational profile, generate occupancy confidence status, reservation deviation status and indoor air quality conflict status, and determine the regional label of the spatial unit; The boundary generation module is used to generate control boundaries and sequence holding conditions based on the area markers, equipment interlocking relationships, manual intervention holding status, and the current collaborative control sequence. The sequence arrangement module is used to write the default control sequence into the current cooperative control sequence when entering automatic control for the first time, and after the minimum hold time is met and the condition of prohibiting reverse switching is lifted, select a cooperative control sequence from the control sequence library that matches the control boundary and the sequence hold condition; otherwise, maintain the current cooperative control sequence. The hierarchical execution module is used to send the collaborative control sequence to the target device according to the hierarchy of the regional controller and the device controller; The verification rollback module is used to perform consistency verification based on execution feedback data, and to switch to a conservative control sequence and update the region marker when cross-source conflicts persist, execution deviations persist, or manual intervention maintains the status.

9. The collaborative energy-saving optimization system based on multi-source data analysis and judgment according to claim 8, characterized in that, It also includes a storage module and a processing module. The storage module stores a space unit mapping table, a device service relationship table, a control sequence library, and a conservative control sequence. The processing module calls the space unit mapping table, the device service relationship table, the control sequence library, and the conservative control sequence to execute the collaborative energy-saving optimization method based on multi-source data analysis and judgment as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the collaborative energy-saving optimization method based on multi-source data analysis and judgment as described in any one of claims 1 to 7.

Images