An emotion-driven multi-zone adaptive optimization intelligent temperature control method for open space

Abstract

This invention discloses an emotion-driven multi-region adaptive joint optimization intelligent temperature control method for open spaces. The method includes establishing a correspondence between regions and temperature control execution units, acquiring region state data, calculating the region's emotional stress index and its corresponding confidence level, establishing an inter-regional influence weight matrix to obtain coupled emotional stress indices for each region and determining influence coefficients, constructing a multi-regional joint optimization model, determining whether emotional event triggering conditions are met, solving the multi-regional joint optimization model to obtain the optimal control vector, generating and issuing control commands, and acquiring environmental parameters and temperature control interaction disturbance parameters after execution in the next control cycle, updating the method based on evaluation results. The beneficial effects of this invention are: it can achieve timely, coordinated, and stable temperature control in open, multi-user shared spaces, improving comfort and reducing fluctuations caused by human intervention.

Description

Technical Field

[0001] This invention relates to the field of intelligent temperature control technology, and in particular to an emotion-driven multi-region adaptive optimization intelligent temperature control method for open spaces. Background Technology

[0002] In existing building or air conditioning temperature control technologies, the common practice is to collect indoor temperature, humidity, air quality indicators, and personnel-related information, calculate the target temperature, select the operating mode, and then adjust the air conditioning operation based on errors. For example, invention application CN116839185A, entitled "An Air Conditioning Temperature Control Method and System," proposes to collect environmental data such as indoor temperature, indoor humidity, indoor VOC concentration, indoor CO2 concentration, indoor occupancy, and outdoor temperature in real time, calculate the target air conditioning temperature, select cooling, heating, and ventilation modes based on the target temperature, indoor CO2 concentration, and outdoor temperature, adjust the air conditioning operation based on error signals, and adjust the air conditioning fan speed based on indoor occupancy.

[0003] Another type of technology tends to combine building structure with functional areas at the building level, first obtaining adjustment data on airflow or temperature distribution in each functional area, and then compensating for changes in pedestrian density. For example, CN119196885A, entitled "An Intelligent Temperature Control Method and System Based on Smart Buildings," uses airflow simulation software to simulate and adjust the airflow path, wind speed, and temperature distribution in each functional area, obtaining temperature-coupled air circulation adjustment data. It also calculates pedestrian density distribution based on real-time monitoring data and then compensates for the indoor temperature circulation adjustment in different functional areas based on the pedestrian density distribution. The compensation algorithm can employ control algorithms such as PID control.

[0004] While the aforementioned existing technologies can adjust temperature control to some extent based on changes in environmental indicators and the number of people, they are still prone to inadequacies in open spaces (such as open office areas and business halls). Firstly, control inputs primarily focus on indicators such as temperature and humidity, pollutant concentration, pedestrian flow, and pedestrian density. Personnel-related information is often used for adjustments like "wind speed adjustment" or "density compensation," making it difficult to characterize the fluctuations in perceived comfort and subjective discomfort caused by queuing, gatherings, and short-term peaks in shared spaces. Consequently, it is difficult to adjust control targets or intensity in a timely manner. Secondly, in open spaces, there is often a transmission of influence between dynamic areas and adjacent areas. For example, people moving between different areas can cause heat load and environmental conditions to affect each other across multiple areas. Existing technologies primarily adjust or compensate for changes in the environment or density of a single area, lacking a framework for describing and processing "cross-area influences that change with pedestrian movement." Therefore, under multi-area coupling conditions, response lag or insufficient coordination in regulation is likely to occur.

[0005] Therefore, the main problems with existing technologies include: how to more timely reflect changes brought about by the flow of people and short-term peaks in open spaces; how to conduct more reasonable zoning adjustments under the condition of mutual influence among multiple areas; and how to reduce frequent adjustments of setpoints and manual intervention while taking into account energy consumption and equipment constraints, so as to improve operational stability and comfort. Summary of the Invention

[0006] To address the problems of existing temperature control technologies that primarily rely on single-area adjustments or compensation based on temperature, humidity, air quality indicators, and pedestrian flow and density, making it difficult to reflect subjective discomfort changes caused by human activity and short-term peaks in open spaces, and lacking cross-area impact characterization under multi-area coupling conditions, resulting in response lag and uncoordinated adjustments, this invention provides an emotion-driven multi-area adaptive joint optimization intelligent temperature control method for open spaces.

[0007] To achieve the above-mentioned objectives, this invention provides an emotion-driven multi-region adaptive joint optimization intelligent temperature control method for open spaces, the method comprising:

[0008] Step S1: Divide the open space into zones, establish the correspondence between each zone and the temperature control execution unit, and periodically acquire the zone status data of each zone; the zone status data includes environmental parameters, equipment operating parameters, acoustic statistical parameters, personnel flow parameters, and temperature control interaction disturbance parameters;

[0009] Step S2: Based on the regional state data, calculate the regional emotional stress index and its corresponding confidence level for each region; the confidence level is used to characterize the reliability of the regional emotional stress index.

[0010] Step S3: Establish an inter-regional influence weight matrix based on personnel flow parameters, and couple the regional emotional stress index of each region according to the inter-regional influence weight matrix to obtain the coupled emotional stress index of each region.

[0011] Step S4: Based on the confidence level of step S2, determine the influence coefficient of the coupled emotional stress index on the optimization solution. The influence coefficient is used to characterize the degree to which the coupled emotional stress index participates in the control optimization.

[0012] Step S5: Based on the environmental parameters, equipment operating parameters, the coupled emotional stress index, and the influence coefficient of each region, construct a multi-region joint optimization model and determine the objective function and constraints of the multi-region joint optimization model;

[0013] Step S6: Determine whether the emotional event triggering condition is met. If the emotional event triggering condition is met, update the event scheduling of the multi-region joint optimization model. If the emotional event triggering condition is not met, maintain the default target weights and constraint parameters.

[0014] Step S7: Based on the target weights and constraint parameters updated or maintained in step S6, solve the multi-region joint optimization model determined in steps S5 and S6 to obtain the optimal control vector for each region, and generate control commands according to the correspondence between each region and the temperature control execution unit, and send the control commands to the corresponding temperature control execution unit for execution.

[0015] Step S8: In the next control cycle, obtain the environmental parameters and temperature control interaction disturbance parameters after execution, generate the control effect evaluation result according to the preset evaluation rules, and update the model parameters used in steps S3 to S7 based on the control effect evaluation result to achieve multi-region adaptive joint optimization control.

[0016] Preferably, in step S1, the area division is determined based on the building plan boundary, the supply and return air coverage area, and the control range of the temperature control execution unit; the correspondence between each area and the temperature control execution unit is one-to-one or one-to-many, and the temperature control execution unit includes at least one of the following: area supply air volume adjustment unit, supply air temperature adjustment unit, or fresh air volume adjustment unit; the environmental parameters include at least one of the following: area air temperature, area relative humidity, and area carbon dioxide concentration; the temperature control interaction disturbance parameters include at least one of the following: number of set value changes, set value change magnitude, and manual coverage duration; the personnel flow parameters include at least one of the following: occupancy of each area and inter-area flow; the occupancy of each area or the inter-area flow is determined based on target point cloud data output by millimeter-wave radar, and the inter-area flow is determined by the number of times the target trajectory crosses the area boundary.

[0017] Preferably, in step S2, all parameters in the regional state data are constructed into regional feature vectors to characterize the corresponding regional state, and the regional feature vectors are normalized. The normalized regional feature vectors are input into the emotional stress inference model, and the output results are smoothed over time to obtain a regional emotional stress index and a category probability distribution corresponding to the regional emotional stress index. A confidence level is derived based on the category probability distribution. The confidence level is used to characterize the reliability of the regional emotional stress index, and the confidence level increases with the concentration of the category probability distribution. The regional emotional stress index is smoothed over time to obtain the regional emotional stress index. The emotional stress inference model is a mapping model trained based on historical sample data, using a lightweight multilayer perceptron, including an input layer, at least two fully connected hidden layers, and an output layer. The hidden layers use a nonlinear activation function to nonlinearly map the input features, and the output layer outputs the category probability distribution corresponding to the emotional stress category. The historical sample data includes regional feature vector samples and their corresponding regional emotional stress labels, which are jointly determined by the changing trends of temperature control interaction perturbation parameters and environmental parameters.

[0018] Preferably, step S3 specifically involves: in the first... t Within each control cycle, based on the statistics of personnel flow parameters, from the first [number] time window... r Each region enters the first s Counting of directional population movement in each region And count the directional personnel flow. Converted to inter-regional flow intensity The inter-regional flow intensity satisfy:

[0019]

[0020] in, The preset time window for statistical analysis of population movement is the same as or an integer multiple of the control period; after obtaining the inter-regional flow intensity between each pair of regions, an inter-regional influence weight matrix is ​​constructed. and the flow intensity between the regions Normalization is performed to obtain the inter-regional influence weight matrix. elements The inter-regional influence weights Indicates the first r Each region for the first s The degree of influence in each region, and the weight of influence between the regions. satisfy:

[0021]

[0022] After obtaining the inter-regional influence weight matrix, the smoothed regional emotional stress index of each region is coupled across regions based on the inter-regional influence weight matrix to obtain the first... s Coupled Emotional Stress Indicators for Each Region And the coupled emotional stress index satisfy:

[0023]

[0024] in, t Indicates the control cycle number. r and s Indicates the area code, and , q This represents the region number used for normalized summation, and M represents the total number of regions; Indicates the duration of the preset time window used for statistical analysis of population movement; Indicates the first t Within the preset time window corresponding to the first control cycle, from the first... The region entered the first Population movement counts for each area; Indicates the first t Within the control cycle, from the first r The area to the first s The intensity of population movement in each region; Indicates the first t Inter-regional influence weight matrix for each control cycle; Indicates the first t Within the first control cycle, the first r The region for the first s The influence weight of each region; Indicates the first r The region in the first t Smoothed regional emotional stress index after one control cycle; Indicates the first s The region in the first t Coupled emotional stress indicators for each control cycle.

[0025] Preferably, step S4 specifically involves: based on the first t Confidence levels of each region within each control period and a set of pre-set reliability threshold parameters Determine the influence coefficient of emotional participation in each region And based on the said emotional participation influence coefficient Adjusting the Coupled Emotional Stress Indicator The degree of participation in the multi-region joint optimization model, wherein the emotional participation influence coefficient is... satisfy:

[0026]

[0027] And satisfy the effective coupling of emotional stress indicators after emotional participation. The calculation relationship is as follows:

[0028]

[0029] in: Indicates the first s The region in the first t Confidence level for each control cycle This indicates the first preset confidence threshold; This indicates the second preset confidence threshold, and satisfies... ; This represents the set of pre-set reliability threshold parameters, including and ; Indicates the first s The region in the first t The emotional participation impact coefficient for each control cycle, with a value range of [value missing]. ; Indicates the first s The region in the first t Coupled emotional stress indicators for each control cycle; Indicates the first s The region in the first t Effective coupling of emotional stress indicators across control cycles.

[0030] Preferably, the multi-region joint optimization model in step S5 uses energy consumption index, comfort deviation index, and control stability index as optimization objectives, and includes constraints on the capacity of the temperature control actuator and the control change rate; wherein,

[0031] The energy consumption index represents the energy consumption of the temperature control execution unit, the comfort deviation index represents the degree of deviation of the regional environmental parameters from the preset comfort target, and the control stability index represents the magnitude or frequency of change of the control quantity.

[0032] The capability constraint of the temperature control actuator is used to limit the output control range of the temperature control actuator, and the control change rate constraint is used to limit the change range of the control quantity between adjacent control cycles.

[0033] Step S6 specifically involves: in the... Within each control cycle, based on at least two of acoustic statistical parameters, personnel flow parameters, temperature control interaction disturbance parameters, and smoothed regional emotional stress index, an emotional event judgment quantity is calculated to characterize the degree of regional state change. When the emotional event judgment quantity reaches a preset event threshold, the corresponding region is determined to meet the emotional event triggering condition, and the multi-region joint optimization model is updated for event scheduling. The event scheduling update includes adjusting at least one of the target weight coefficient and constraint parameters of the multi-region joint optimization model. When the emotional event judgment quantity does not reach the preset event threshold, the corresponding region is determined not to meet the emotional event triggering condition, and the default target weight and constraint parameters are maintained.

[0034] The beneficial effects of this invention are as follows: In an open, multi-user shared space, this invention achieves more timely, coordinated, and stable temperature control for different zones, improving comfort and reducing fluctuations caused by human intervention while considering energy consumption and equipment operational constraints. Specifically, compared to existing technologies that primarily rely on temperature, humidity, air quality, and pedestrian flow or density for adjustment or compensation, this invention can more effectively reflect the impact of changes in human activity and short-term peaks on perceived fluctuations and subjective discomfort, thereby improving the timeliness of control target and intensity adjustments. In scenarios where multiple zones significantly influence each other, this invention can more reasonably characterize the influence relationships between zones as people move, making zoned adjustments more coordinated and reducing response lag and local over- or under-adjustment. Simultaneously, by incorporating comprehensive consideration of operational stability and equipment constraints during the control process, this invention reduces the probability of frequent setpoint adjustments and human intervention, minimizes instability caused by drastic changes in control parameters, and thus improves the long-term stability of the system and the consistency of user experience. Attached Figure Description

[0035] Figure 1 This is a flowchart of the method according to Embodiment 1 of the present invention. Detailed Implementation

[0036] To clearly illustrate the technical features of this solution, the following detailed implementation method will be used to explain the solution.

[0037] Example 1

[0038] See Figure 1 This invention provides an emotion-driven multi-region adaptive joint optimization intelligent temperature control method for open spaces, the method comprising:

[0039] Step S1: Divide the open space into zones, establish the correspondence between each zone and the temperature control execution unit, and periodically acquire the zone status data of each zone; the zone status data includes environmental parameters, equipment operating parameters, acoustic statistical parameters, personnel flow parameters, and temperature control interaction disturbance parameters;

[0040] Step S2: Based on the regional status data, calculate the regional emotional stress index for each region and its corresponding confidence level; the confidence level is used to characterize the reliability of the regional emotional stress index.

[0041] Step S3: Establish an inter-regional influence weight matrix based on personnel flow parameters, and couple the regional emotional stress index of each region according to the inter-regional influence weight matrix to obtain the coupled emotional stress index of each region.

[0042] Step S4: Based on the confidence level in step S2, determine the influence coefficient of the coupled emotional stress index on the optimization solution. The influence coefficient is used to characterize the degree to which the coupled emotional stress index participates in the control optimization.

[0043] Step S5: Based on the environmental parameters, equipment operating parameters, coupled emotional stress index and influence coefficient of each region, construct a multi-region joint optimization model and determine the objective function and constraints of the multi-region joint optimization model;

[0044] Step S6: Determine whether the emotional event triggering conditions are met. If the emotional event triggering conditions are met, update the multi-region joint optimization model by scheduling events. If the emotional event triggering conditions are not met, maintain the default target weights and constraint parameters.

[0045] Step S7: Based on the target weights and constraint parameters updated or maintained in step S6, solve the multi-region joint optimization model determined in steps S5 and S6 to obtain the optimal control vector for each region, and generate control commands according to the correspondence between each region and the temperature control execution unit, and send the control commands to the corresponding temperature control execution unit for execution.

[0046] Step S8: In the next control cycle, obtain the environmental parameters and temperature control interaction disturbance parameters after execution, generate the control effect evaluation result according to the preset evaluation rules, and update the model parameters used in steps S3 to S7 based on the control effect evaluation result to achieve multi-region adaptive joint optimization control.

[0047] Preferably, in step S1, the area division is determined based on the building plan boundary, the supply and return air coverage area, and the control range of the temperature control execution unit; the correspondence between each area and the temperature control execution unit is one-to-one or one-to-many, and the temperature control execution unit includes at least one of the following: area supply air volume adjustment unit, supply air temperature adjustment unit, or fresh air volume adjustment unit; environmental parameters include at least one of the following: area air temperature, area relative humidity, and area carbon dioxide concentration; temperature control interaction disturbance parameters include at least one of the following: number of set value changes, set value change magnitude, and manual coverage duration; personnel flow parameters include at least one of the following: occupancy of each area and inter-area flow; the occupancy of each area or inter-area flow is determined based on the target point cloud data output by millimeter-wave radar, and the inter-area flow is determined by the number of times the target trajectory crosses the area boundary.

[0048] In step S2, all parameters in the regional state data are constructed into regional feature vectors to represent the corresponding regional states, and these feature vectors are normalized. The normalized regional feature vectors are then input into the emotional stress inference model. The output is smoothed over time to obtain the regional emotional stress index and the corresponding category probability distribution. A confidence level is derived based on the category probability distribution. The confidence level characterizes the reliability of the regional emotional stress index, and it increases with the concentration of the category probability distribution. The regional emotional stress index is then smoothed over time to obtain the regional emotional stress index. The emotional stress inference model is a mapping model trained on historical sample data. It employs a lightweight multilayer perceptron, including an input layer, at least two fully connected hidden layers, and an output layer. The hidden layers use nonlinear activation functions to nonlinearly map the input features, and the output layer outputs the category probability distribution corresponding to the emotional stress category. The historical sample data includes regional feature vector samples and their corresponding regional emotional stress labels. The regional emotional stress labels are jointly determined by the changing trends of temperature-controlled interaction perturbation parameters and environmental parameters.

[0049] Step S3 specifically involves: in the... t Within each control cycle, based on the statistics of personnel flow parameters, from the first [number] time window... r Each region enters the first s Counting of directional population movement in each region And count directional personnel movement. Converted to inter-regional flow intensity Interregional flow intensity satisfy:

[0050]

[0051] in, The preset time window for statistical analysis of population movement is the same as or an integer multiple of the control period; after obtaining the inter-regional flow intensity between each pair of regions, an inter-regional influence weight matrix is ​​constructed. and the inter-regional flow intensity Normalization is performed to obtain the inter-regional influence weight matrix. elements The weight of regional influence Indicates the first r Each region for the first s The degree of influence in each region, and the weight of influence between regions. satisfy:

[0052]

[0053] After obtaining the inter-regional influence weight matrix, the smoothed regional emotional stress index of each region is coupled across regions based on the inter-regional influence weight matrix to obtain the first... s Coupled Emotional Stress Indicators for Each Region And coupled with emotional stress indicators satisfy:

[0054]

[0055] in, t Indicates the control cycle number. r and s Indicates the area code, and , q This represents the region number used for normalized summation, and M represents the total number of regions; Indicates the duration of the preset time window used for statistical analysis of population movement; Indicates the first t Within the preset time window corresponding to the first control cycle, from the first... The region entered the first Population movement counts for each area; Indicates the first t Within the control cycle, from the first r The area to the first s The intensity of population movement in each region; Indicates the first t Inter-regional influence weight matrix for each control cycle; Indicates the first t Within the first control cycle, the first r The region for the first s The influence weight of each region; Indicates the first r The region in the first t Smoothed regional emotional stress index after one control cycle; Indicates the first s The region in the first t Coupled emotional stress indicators for each control cycle.

[0056] Step S4 specifically involves: based on the first t Confidence levels of each region within each control period and a set of pre-set reliability threshold parameters Determine the influence coefficient of emotional participation in each region And based on the influence coefficient of emotional participation Adjusting the Coupled Emotional Stress Indicator The degree of participation in the multi-region joint optimization model, including the influence coefficient of emotional participation. satisfy:

[0057]

[0058] And satisfy the effective coupling of emotional stress indicators after emotional participation. The calculation relationship is as follows:

[0059]

[0060] in: Indicates the first s The region in the first t Confidence level for each control cycle This indicates the first preset confidence threshold; This indicates the second preset confidence threshold, and satisfies... ; This represents the set of pre-set reliability threshold parameters, including and ; Indicates the first s The region in the first t The emotional participation impact coefficient for each control cycle, with a value range of [value missing]. ; Indicates the first s The region in the first t Coupled emotional stress indicators for each control cycle; Indicates the first s The region in the first t Effective coupling of emotional stress indicators across control cycles.

[0061] The multi-region joint optimization model in step S5 uses energy consumption index, comfort deviation index, and control stability index as optimization objectives, and includes temperature control actuator capability constraints and control change rate constraints. Among them, the energy consumption index represents the energy consumption of the temperature control actuator, the comfort deviation index represents the degree of deviation of the regional environmental parameters from the preset comfort target, and the control stability index represents the magnitude or frequency of change of the control quantity. The temperature control actuator capability constraint is used to limit the output control range of the temperature control actuator, and the control change rate constraint is used to limit the magnitude of change of the control quantity between adjacent control cycles.

[0062] Step S6 specifically involves: in the... Within each control cycle, based on at least two of the acoustic statistical parameters, personnel flow parameters, temperature control interaction disturbance parameters, and smoothed regional emotional stress index, an emotional event judgment quantity is calculated to characterize the degree of regional state change. When the emotional event judgment quantity reaches a preset event threshold, the corresponding region is determined to meet the emotional event triggering condition, and the multi-region joint optimization model is updated by event scheduling. The event scheduling update includes adjusting at least one of the target weight coefficient and constraint parameters of the multi-region joint optimization model. When the emotional event judgment quantity does not reach the preset event threshold, the corresponding region is determined not to meet the emotional event triggering condition, and the default target weight and constraint parameters are maintained.

[0063] Example 2

[0064] This invention provides an emotion-driven, multi-region adaptive joint optimization intelligent temperature control method for open spaces. This method is based on the characteristics of open, shared spaces with multiple users, including "the same air conditioning system serving multiple workstations, dynamic changes in personnel status and area load, and the ease with which control commands can couple and interfere." It introduces emotional stress as an auxiliary adjustment quantity reflecting changes in group thermal comfort needs into the joint optimization framework of multi-region temperature control. Specifically, the open space is first divided into regions, and a correspondence between regions and temperature control execution units is established. Regional state data, such as environmental conditions, equipment operation, acoustic statistics, millimeter-wave point cloud data on pedestrian flow, and temperature control interaction disturbances, are periodically collected. Based on this, a regional emotional stress index is inferred for each region, and a confidence level is given to measure the reliability of the emotional stress estimation. Subsequently, an inter-regional influence weight matrix is ​​constructed using cross-regional personnel flow information. Regional emotional stress is then coupled across regions to obtain a coupled emotional stress index that reflects the impact of personnel flow. Based on the confidence level, its influence coefficient in control optimization is determined, thereby avoiding misleading control by low-reliability emotional signals. Furthermore, a multi-region joint optimization model is constructed with energy consumption, comfort deviation, and control stability as objectives, and constrained by the execution unit capability and control change rate. When sudden changes occur in acoustics, pedestrian flow, interactive disturbances, or emotional stress, emotional event scheduling is triggered, dynamically updating the objective weights and constraint parameters of the optimization model. Finally, in each control cycle, the control quantities for each region are solved and executed. In the next control cycle, a control effect evaluation result is formed based on environmental feedback and interactive disturbances, thereby adaptively updating the coupling weights, confidence gating parameters, and optimization model parameters, realizing an emotion-driven multi-region adaptive joint optimization temperature control closed loop for open spaces. Specifically, this method includes the following steps:

[0065] 1. Divide the open space into zones, establish the correspondence between each zone and the temperature control actuator, and periodically acquire the zone status data for each zone. Zone status data includes environmental parameters, equipment operating parameters, acoustic statistical parameters, personnel flow parameters, and temperature control interaction disturbance parameters. This zone division is based on the building plan boundaries, supply and return air coverage, and the control range of the temperature control actuator. The correspondence between each zone and the temperature control actuator is one-to-one or one-to-many, and the temperature control actuator includes at least one of the following: zone air volume adjustment unit, supply air temperature adjustment unit, or fresh air volume adjustment unit.

[0066] Environmental parameters include regional air temperature, regional relative humidity, and regional carbon dioxide concentration; temperature control interaction disturbance parameters include the number of times the set value is changed, the magnitude of the set value change, and the duration of manual coverage; personnel flow parameters include the occupancy of each area and the flow between areas; the occupancy of each area or the flow between areas is determined based on the target point cloud data output by millimeter-wave radar, and the flow between areas is determined by the number of times the target trajectory crosses the area boundary.

[0067] Second, all parameters in the regional state data are constructed into regional feature vectors to characterize the corresponding regional state, and these feature vectors are normalized. Then, the normalized regional feature vectors are input into the emotional stress inference model. The output results are smoothed over time to obtain the regional emotional stress index and the corresponding category probability distribution. A confidence score is derived based on the category probability distribution. The confidence score characterizes the reliability of the regional emotional stress index, and it increases with the concentration of the category probability distribution. The regional emotional stress index is then smoothed over time to obtain the final regional emotional stress index.

[0068] Specifically, in this embodiment, the emotional stress inference model is a mapping model trained based on historical sample data. It adopts a lightweight multilayer perceptron, including an input layer, at least two fully connected hidden layers, and an output layer. The hidden layers use non-linear activation functions to perform non-linear mapping on the input features, and the output layer is used to output the category probability distribution corresponding to the emotional stress category.

[0069] Historical sample data includes regional feature vector samples and their corresponding regional emotional stress labels. The regional emotional stress labels are jointly determined by the changing trends of temperature control interaction perturbation parameters and environmental parameters.

[0070] Third, establish an inter-regional influence weight matrix based on personnel flow parameters, and couple the regional emotional stress index of each region according to the inter-regional influence weight matrix to obtain the coupled emotional stress index of each region. This step specifically involves: in the first... t Within each control cycle, based on the statistics of personnel flow parameters, from the first [number] time window... r Each region enters the first s Counting of directional population movement in each region And count directional personnel movement. Converted to inter-regional flow intensity Interregional flow intensity satisfy:

[0071]

[0072] in, The preset time window for statistical analysis of population movement is the same as or an integer multiple of the control period; after obtaining the inter-regional flow intensity between each pair of regions, an inter-regional influence weight matrix is ​​constructed. and the inter-regional flow intensity Normalization is performed to obtain the inter-regional influence weight matrix. elements The weight of regional influence Indicates the firstr Each region for the first s The degree of influence in each region, and the weight of influence between regions. satisfy:

[0073]

[0074] After obtaining the inter-regional influence weight matrix, the smoothed regional emotional stress index of each region is coupled across regions based on the inter-regional influence weight matrix to obtain the first... s Coupled Emotional Stress Indicators for Each Region And coupled with emotional stress indicators satisfy:

[0075]

[0076] in, t Indicates the control cycle number. r and s Indicates the area code, and , q This represents the region number used for normalized summation, and , M Indicates the total number of regions; Indicates the duration of the preset time window used for statistical analysis of population movement; Indicates the first t Within the preset time window corresponding to the first control cycle, from the first... The region entered the first Population movement counts for each area; Indicates the first t Within the control cycle, from the first r The area to the first s The intensity of population movement in each region; Indicates the first t Inter-regional influence weight matrix for each control cycle; Indicates the first t Within the first control cycle, the first r The region for the first s The influence weight of each region; Indicates the first r The region in the first t Smoothed regional emotional stress index after one control cycle; Indicates the first s The region in the first t Coupled emotional stress indicators for each control cycle.

[0077] Fourth, based on the confidence level in step two, determine the influence coefficient of the coupled emotional stress index on the optimization solution. The influence coefficient is used to characterize the degree to which the coupled emotional stress index participates in controlling the optimization. Specifically, this step involves: based on the confidence level in step two...t Confidence levels of each region within each control period and a set of pre-set reliability threshold parameters Determine the influence coefficient of emotional participation in each region And based on the influence coefficient of emotional participation Adjusting the Coupled Emotional Stress Indicator The degree of participation in the multi-region joint optimization model, including the influence coefficient of emotional participation. satisfy:

[0078]

[0079] And satisfy the effective coupling of emotional stress indicators after emotional participation. The calculation relationship is as follows:

[0080]

[0081] in: Indicates the first s The region in the first t Confidence level for each control cycle This indicates the first preset confidence threshold; This indicates the second preset confidence threshold, and satisfies... ; This represents the set of pre-set reliability threshold parameters, including and ; Indicates the first s The region in the first t The emotional participation impact coefficient for each control cycle, with a value range of [value missing]. ; Indicates the first s The region in the first t Coupled emotional stress indicators for each control cycle; Indicates the first s The region in the first t Effective coupling of emotional stress indicators across control cycles.

[0082] V. Based on environmental parameters, equipment operating parameters, coupled emotional stress indicators, and influence coefficients of each region, a multi-regional joint optimization model is constructed. This model uses energy consumption, comfort deviation, and control stability indicators as optimization objectives, and includes constraints on the temperature control actuator's capacity and the rate of change of control.

[0083] Energy consumption index represents the energy consumption of the temperature control actuator; comfort deviation index represents the degree of deviation of the regional environmental parameters from the preset comfort target; control stability index represents the magnitude or frequency of change of the control quantity.

[0084] The temperature control actuator capability constraint is used to limit the output control range of the temperature control actuator, and the control change rate constraint is used to limit the change range of the control quantity between adjacent control cycles.

[0085] The multi-region joint optimization model is as follows:

[0086] With the first t Control vector of each region within each control cycle As decision variables, with the objective function Minimization is the optimization objective, and the objective function and constraints of the multi-region joint optimization model are determined under the conditions of satisfying the capacity constraints of the temperature control actuator and the control rate of change constraints; where the objective function is... It consists of energy consumption indicators, comfort deviation indicators, and control stability indicators, and satisfies the following:

[0087]

[0088] The temperature control actuator capability constraint limits the output control range of the control vector, and the control change rate constraint limits the change amplitude of the control vector between adjacent control cycles, and respectively satisfy the following:

[0089]

[0090]

[0091] And comfort deviation index Used to indicate the degree of deviation of regional environmental parameters from preset comfort targets, controlling stability indicators. Used to represent the magnitude or frequency of change of a control quantity, and respectively satisfying:

[0092] ,

[0093] in, Indicates the first s The region in the first t The control quantity vector for each control cycle; Indicates the first s The region in the first t -1 control quantity vector for a control cycle; , They represent the first s The lower and upper bound vectors of the control vector for each region; This represents the preset weighting coefficient corresponding to the comfort deviation index; This represents the preset weighting coefficient corresponding to the energy consumption index; Indicates the first sThe vector representing the maximum permissible change in the control quantity vector of a region between adjacent control cycles.

[0094] VI. In the Within each control cycle, based on at least two of the acoustic statistical parameters, personnel flow parameters, temperature control interaction disturbance parameters, and smoothed regional emotional stress index, an emotional event judgment quantity is calculated to characterize the degree of regional state abrupt change. When the emotional event judgment quantity reaches a preset event threshold, the corresponding region is determined to meet the emotional event triggering condition, and the multi-region joint optimization model is updated through event scheduling. The event scheduling update includes adjusting at least one of the target weight coefficient and constraint parameters of the multi-region joint optimization model. When the emotional event judgment quantity does not reach the preset event threshold, the corresponding region is determined not to meet the emotional event triggering condition, and the default target weight and constraint parameters are maintained. Details:

[0095] The triggering conditions for emotional events are determined by the emotional event judgment quantity. With preset event threshold It is determined that, within the t-th control period, the acoustic abrupt change degree is calculated based on acoustic statistical parameters, personnel flow parameters, and temperature control interaction disturbance parameters, respectively. Fluid abrupt change and interactive mutation degree And based on acoustic abrupt change Fluid abrupt change Interactive mutation degree and the smoothed regional emotional stress index At least two of them are used to construct the emotional event judgment metric. ;

[0096] when At that time, the judgment of the first s Each region satisfies the emotional event triggering condition, and the objective weight coefficients and constraint parameters of the multi-region joint optimization model are updated through event scheduling. Among these, the emotional event judgment quantity... satisfy:

[0097]

[0098] Acoustic abrupt change satisfy:

[0099]

[0100] Fluid abrupt change satisfy:

[0101]

[0102] Interactive mutation degree satisfy:

[0103]

[0104] Furthermore, when the emotional event triggering conditions are met, the event scheduling update of the multi-region joint optimization model includes at least one of the following: increasing the preset weight coefficient corresponding to the comfort deviation index. Increase the maximum allowable change vector corresponding to the control rate of change constraint. And adjust the preset weighting coefficients corresponding to the energy consumption indicators. ;

[0105] in, Indicates the first s The region in the first t The amount of emotional events judged in each control cycle; These represent the preset fusion weight coefficients corresponding to acoustic abruptness, fluid abruptness, interaction abruptness, and smoothed regional emotional stress index, respectively, and all are non-negative numbers;

[0106] Indicates the first s The region in the first t Acoustic abrupt change in each control cycle; Indicates the first s The region in the first t The degree of flow abrupt change in each control cycle; Indicates the first s The region in the first t The degree of interactive mutation in each control cycle; Indicates the first s The region in the first t The periodic mean of acoustic statistics for each control cycle; Indicates the first s The region in the first t -1 control cycle acoustic statistics periodic mean; Indicates the first s The region in the first t The total amount of personnel mobility intensity for each control cycle; Indicates the first s The region in the first t -The total amount of personnel mobility intensity over one control period; Indicates the s-th region in the th order. t The sum of the temperature control interactive disturbance intensity for each control cycle; Indicates the first s The region in the first t -The sum of the temperature control interactive disturbance intensity over one control cycle; Indicates the first The region in the first The smoothed regional emotional stress index after a control cycle.

[0107] VII. Based on the target weights and constraint parameters from step VI, solve the multi-region joint optimization model determined in steps VI and VII to obtain the optimal control vector for each region. Then, generate control commands according to the correspondence between each region and the temperature control execution unit, and issue the control commands to the corresponding temperature control execution unit for execution. Specifically, in the… t Within each control cycle, based on the target weight coefficients and constraint parameters updated or maintained in step six, the multi-region joint optimization model constructed in step five is iteratively solved to obtain the optimal control vector for each region. Based on the correspondence between each region and the temperature control execution unit established in step one, the optimal control vector is... Converted into a set of control commands sent to the temperature control actuator. The transformation satisfies:

[0108]

[0109] And for the optimal control vector The iterative solution uses a stopping criterion. The iteration terminates and the optimal control vector is output when the absolute value of the difference between the objective function values ​​obtained from two consecutive iterations is less than the stopping criterion. The stopping criterion satisfies the following:

[0110]

[0111] in, Indicates the first s The region in the first t The optimal control vector for each control cycle; Indicates the first s The mapping function from the control vector of each region to the control command is determined by the correspondence between the region and the temperature control execution unit; Represents the set union operation Indicates the first t Within the first control cycle, the first j The objective function value obtained in the next iteration; Indicates the first t Within the first control cycle, the first j The objective function value obtained in one iteration; j Indicates the iteration number; This represents the preset stopping criterion threshold for iterative solution.

[0112] 8. In the next control cycle, acquire the environmental parameters and temperature control interaction disturbance parameters after execution, generate control effect evaluation results according to preset evaluation rules, and update the model parameters used in steps three to seven based on the control effect evaluation results to achieve multi-region adaptive joint optimization control. Details: In the... t+1 control cycle to obtain feedback data after executing step seven. The feedback data shall include at least the environmental parameter vectors of each region. and the total amount of temperature control interaction disturbance parameters ; Calculate the control effectiveness evaluation value for each region based on feedback data. And evaluate the quantity based on the control effect. Generate control effect evaluation results; including control effect evaluation quantities Used to characterize the combined changing trend of comfort bias, control stability, and interactive disturbances within adjacent control cycles, and satisfying the following:

[0113]

[0114] in and represent the improvement of the comfort deviation index, control stability index, and temperature control interaction disturbance intensity in adjacent control cycles, respectively, and satisfy the following:

[0115]

[0116]

[0117]

[0118] in, Indicates the first s The region in the first t +1 control cycle of environmental parameter vector; and They represent the first s The region in the first t The first control cycle and the first t The sum of temperature control interactive disturbance parameters for +1 control cycle; and They represent the first s The region in the first t The first control cycle and the first t Comfort deviation index for +1 control cycle; and They represent the first s The region in the first t The first control cycle and the first t Control stability index for +1 control cycle; Indicates the first s The improvement in comfort deviation indicators in each region within adjacent control periods; Indicates the first s The amount of improvement in control stability indicators for each region within adjacent control periods; Indicates the first s The improvement in the intensity of temperature control interaction disturbance in each region within adjacent control cycles; Indicates the first s The region in the first t +1 control effect evaluation quantity for control cycle; These represent the preset evaluation weight coefficients corresponding to the improvement in comfort deviation index, the improvement in control stability index, and the improvement in interaction disturbance, respectively, and all are non-negative numbers; A set of threshold parameters representing the influence coefficient of emotional participation; These represent the preset target weighting coefficients for energy consumption indicators, comfort deviation indicators, and control stability indicators, respectively.

[0119] Then evaluate the control effect based on the quantity. The generated control effect evaluation results include: when When the control effect evaluation result is determined to be an improvement in control effect, The control effect evaluation result is determined to be a deterioration in control effect. Based on the control effect evaluation result, the model parameters used in steps three to seven are adaptively updated to ensure that the inter-regional coupling, confidence gating, event scheduling updates, and multi-regional joint optimization solutions in subsequent control cycles match the control effect evaluation result. The adaptive update includes at least one of the following: updating the calculation parameters of the inter-regional influence weight matrix, and updating the threshold parameter set of the emotion participation influence coefficient. Update, or adjust the preset target weight coefficients of the multi-region joint optimization model. Update.

[0120] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An emotion-driven, multi-region adaptive joint optimization intelligent temperature control method for open spaces, characterized in that: The method includes: Step S1: Divide the open space into zones, establish the correspondence between each zone and the temperature control execution unit, and periodically acquire the zone status data of each zone; the zone status data includes environmental parameters, equipment operating parameters, acoustic statistical parameters, personnel flow parameters, and temperature control interaction disturbance parameters; Step S2: Based on the regional state data, calculate the regional emotional stress index and its corresponding confidence level for each region; Step S3: Establish an inter-regional influence weight matrix based on personnel flow parameters, and couple the regional emotional stress index of each region according to the inter-regional influence weight matrix to obtain the coupled emotional stress index of each region. Step S4: Based on the confidence level of step S2, determine the influence coefficient of the coupled emotional stress index on the optimization solution; Step S5: Based on the environmental parameters, equipment operating parameters, the coupled emotional stress index, and the influence coefficient of each region, construct a multi-region joint optimization model including the objective function and constraints; Step S6: Determine whether the emotional event triggering condition is met. If the emotional event triggering condition is met, update the event scheduling of the multi-region joint optimization model. If the emotional event triggering condition is not met, maintain the default target weights and constraint parameters. Step S7: Based on the target weights and constraint parameters updated or maintained in step S6, solve the multi-region joint optimization model determined in steps S5 and S6 to obtain the optimal control vector for each region, and generate control commands according to the correspondence between each region and the temperature control execution unit, and send the control commands to the corresponding temperature control execution unit for execution. Step S8: In the next control cycle, obtain the environmental parameters and temperature control interaction disturbance parameters after execution, generate the control effect evaluation result according to the preset evaluation rules, and update the model parameters used in steps S3 to S7 based on the control effect evaluation result to achieve multi-region adaptive joint optimization control. The zoning is determined based on the building plan boundary, the supply and return air coverage area, and the control range of the temperature control execution unit; the correspondence between each zone and the temperature control execution unit is one-to-one or one-to-many, and the temperature control execution unit includes at least one of the following: zone supply air volume adjustment unit, supply air temperature adjustment unit, or fresh air volume adjustment unit. The environmental parameters include at least one of the regional air temperature, regional relative humidity, and regional carbon dioxide concentration. The temperature control interactive disturbance parameters include at least one of the following: the number of times the set value is changed, the magnitude of the set value change, and the duration of manual overwrite. The personnel flow parameters include at least one of the occupancy of each area and the flow between areas; the occupancy of each area or the flow between areas is determined based on the target point cloud data output by millimeter-wave radar, and the flow between areas is determined by the number of times the target trajectory crosses the area boundary.

2. The method according to claim 1, characterized in that, In step S2, all parameters in the regional state data are constructed into a regional feature vector to characterize the corresponding regional state, and the regional feature vector is normalized. The normalized regional feature vector is input into the emotional stress inference model. The output is smoothed over time to obtain the regional emotional stress index and the category probability distribution corresponding to the regional emotional stress index. The confidence level is then derived based on the category probability distribution.

3. The method according to claim 2, characterized in that, The emotional stress inference model is a mapping model trained based on historical sample data. It adopts a lightweight multilayer perceptron, including an input layer, at least two fully connected hidden layers, and an output layer. The hidden layers use a non-linear activation function to perform non-linear mapping on the input features, and the output layer is used to output the category probability distribution corresponding to the emotional stress category. The historical sample data includes regional feature vector samples and their corresponding regional emotional stress labels, which are determined by the combined changing trends of temperature control interaction perturbation parameters and environmental parameters.

4. The method according to claim 1, characterized in that, Step S3 specifically involves: in the... t Within each control cycle, based on the statistics of personnel flow parameters, from the first [number] time window... r Each region enters the first s Counting of directional population movement in each region And count the directional personnel flow. Converted to interregional flow intensity The inter-regional flow intensity satisfy: in, The preset time window for statistical analysis of population movement is the same as or an integer multiple of the control period; after obtaining the inter-regional flow intensity between each pair of regions, an inter-regional influence weight matrix is ​​constructed. and the flow intensity between the regions Normalization is performed to obtain the inter-regional influence weight matrix. elements The inter-regional influence weights Indicates the first r Each region for the first s The degree of influence in each region, and the weight of influence between the regions. satisfy: After obtaining the inter-regional influence weight matrix, the smoothed regional emotional stress index of each region is coupled across regions based on the inter-regional influence weight matrix to obtain the first... s Coupled Emotional Stress Indicators for Each Region And the coupled emotional stress index satisfy: in, t Indicates the control cycle number. r and s Indicates the area code, and , q This represents the region number used for normalized summation, and , M Indicates the total number of regions; Indicates the duration of the preset time window used for statistical analysis of population movement; Indicates the first t Within the preset time window corresponding to the first control cycle, from the first... r The region entered the first Population movement counts for each area; Indicates the first t Within the control cycle, from the first r The area to the first The intensity of population movement in each region; Indicates the first t Inter-regional influence weight matrix for each control cycle; Indicates the first t Within the first control cycle, the first r The region for the first s The influence weight of each region; Indicates the first r The region in the first t Smoothed regional emotional stress index after one control cycle; Indicates the first s The region in the first t Coupled emotional stress indicators for each control cycle.

5. The method according to claim 1, characterized in that, Step S4 specifically involves: based on the first t Confidence levels of each region within each control period and a set of pre-set reliability threshold parameters Determine the influence coefficient of emotional participation in each region And based on the said emotional participation influence coefficient Adjusting the Coupled Emotional Stress Indicator The degree of participation in the multi-region joint optimization model, wherein the emotional participation influence coefficient is... satisfy: And satisfy the effective coupling of emotional stress indicators after emotional participation. The calculation relationship is as follows: in: Indicates the first s The region in the first t Confidence level for each control cycle This indicates the first preset confidence threshold; This indicates the second preset confidence threshold, and satisfies... ; This represents the set of pre-set reliability threshold parameters, including and ; Indicates the first s The region in the first t The emotional participation impact coefficient for each control cycle, with a value range of [value missing]. ; Indicates the first s The region in the first t Coupled emotional stress indicators for each control cycle; Indicates the first s The region in the first t Effective coupling of emotional stress indicators across control cycles.

6. The method according to claim 1, characterized in that, The multi-region joint optimization model in step S5 uses energy consumption, comfort deviation, and control stability as optimization objectives, and includes constraints on the temperature control actuator's capability and the rate of change of control; wherein, The energy consumption index represents the energy consumption of the temperature control execution unit, the comfort deviation index represents the degree of deviation of the regional environmental parameters from the preset comfort target, and the control stability index represents the magnitude or frequency of change of the control quantity. The capability constraint of the temperature control actuator is used to limit the output control range of the temperature control actuator, and the control change rate constraint is used to limit the change range of the control quantity between adjacent control cycles.

7. The method according to claim 1, characterized in that, Step S6 specifically involves: in the... Within each control cycle, based on at least two of the acoustic statistical parameters, personnel flow parameters, temperature control interaction disturbance parameters, and smoothed regional emotional stress index, an emotional event judgment quantity is calculated to characterize the degree of regional state change. When the number of emotional events determined reaches the preset event threshold, the corresponding region is determined to meet the emotional event triggering condition, and the multi-region joint optimization model is updated by event scheduling. The event scheduling update includes adjusting at least one of the target weight coefficient and constraint parameters of the multi-region joint optimization model. When the number of emotional events determined does not reach the preset event threshold, the corresponding region is determined not to meet the emotional event triggering condition, and the default target weight and constraint parameters are maintained.

Images