Intelligent temperature adaptive adjusting system based on Internet of Things temperature threshold analysis

Abstract

The invention discloses an intelligent temperature adaptive adjustment system based on Internet of Things temperature threshold analysis, and belongs to the technical field of temperature adjustment, and the system comprises a data sensing transmission module which is used for collecting environment temperature data in real time and transmitting the environment temperature data to a cloud platform; the threshold value dynamic adjustment module is used for dynamically adjusting or recommending threshold values and intelligently managing the threshold values; the threshold value analysis and judgment module is used for judging whether the current environment temperature data exceed a recommended temperature threshold value range or not and determining a threshold value analysis and judgment result; and the coordination control adjustment module is used for adopting different adjustment intensities according to the threshold analysis and judgment result. The intelligent temperature control system solves the problems that intelligent accurate control over the temperature cannot be achieved through real-time monitoring, data analysis and intelligent regulation and control, and comfortable, energy-saving and efficient self-adaptive management cannot be achieved in the prior art. According to the invention, intelligent accurate control of the temperature can be realized through real-time monitoring, data analysis and intelligent regulation and control, and comfortable, energy-saving and efficient adaptive management cannot be realized.

Description

Technical Field

[0001] This invention relates to the field of temperature regulation technology, specifically to an intelligent temperature adaptive regulation system based on Internet of Things (IoT) temperature threshold analysis. Background Technology

[0002] Ambient temperature regulation is an indispensable part of modern life. Its goal is to achieve precise control of ambient temperature through technological means, thereby improving comfort, ensuring equipment operating efficiency, and promoting energy conservation and environmental protection. It is widely used in various scenarios such as homes, industries, agriculture, and logistics.

[0003] Chinese patent application CN114440421B discloses an environmental temperature regulation method and system. The method includes the following steps: obtaining basic parameters of a temperature control device and acquiring user adjustment data; calculating the individual's desired temperature and optimal temperature based on the adjustment data; adjusting the environmental temperature according to the optimal temperature; and performing local temperature control on the user's area based on the individual's desired temperature, location data, and the basic parameters of the temperature control device. This allows for meeting the temperature needs of all indoor occupants while further satisfying the individual's temperature requirements. However, this patent has the following drawbacks:

[0004] Existing technologies cannot achieve intelligent and precise temperature control through real-time monitoring, data analysis, and intelligent regulation, nor can they achieve comfortable, energy-saving, and efficient adaptive management. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent temperature adaptive regulation system based on IoT temperature threshold analysis, which can achieve intelligent and precise temperature control through real-time monitoring, data analysis and intelligent regulation, and can achieve comfortable, energy-saving and efficient adaptive management, thus solving the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] An intelligent temperature adaptive regulation system based on IoT temperature threshold analysis includes:

[0008] The data sensing and transmission module is used to collect ambient temperature data in real time and transmit the collected ambient temperature data to the cloud platform via a wireless network.

[0009] The threshold dynamic adjustment module is used to dynamically adjust or recommend thresholds based on historical data, weather forecasts, user behavior patterns, and energy consumption targets, for intelligent management of thresholds.

[0010] The threshold analysis and judgment module is used to analyze the ambient temperature data and determine whether the current ambient temperature data exceeds the recommended temperature threshold range, and to determine the threshold analysis and judgment result.

[0011] The coordination control and adjustment module is used to adopt different adjustment intensities based on the threshold analysis results, so as to realize multi-device collaborative control and conflict coordination, and ensure intelligent temperature adaptive adjustment.

[0012] Preferably, the threshold is dynamically adjusted or recommended based on historical data, weather forecasts, user behavior patterns, and energy consumption targets, and the following operations are performed:

[0013] Collect and process multi-factor, multi-source data to determine multi-factor fusion data;

[0014] A multi-factor threshold recommendation model is constructed using machine learning algorithms, and the multi-factor threshold recommendation model is deployed in a real multi-factor threshold recommendation environment.

[0015] Multi-factor fusion data is input into a multi-factor threshold recommendation model, and the multi-factor fusion data is analyzed based on the multi-factor threshold recommendation model. The threshold is dynamically adjusted or recommended based on historical data, weather forecasts, user behavior patterns and energy consumption targets.

[0016] By analyzing users' daily activity time, the system automatically switches the corresponding scene thresholds, analyzes the temperature change patterns of the same period over the past 30 days, predicts temperature fluctuations and adjusts the thresholds in advance, and dynamically corrects the indoor thresholds by combining the outdoor temperature and light intensity for the next 24 hours. When the real-time energy consumption approaches the preset monthly and annual targets, the threshold range is automatically tightened to determine the recommended temperature threshold, thus achieving intelligent management of temperature thresholds.

[0017] Preferably, multi-factor, multi-source data is collected and processed to determine multi-factor fusion data, and the following operations are performed:

[0018] It monitors real-time temperature data, user-preset thresholds, historical operating data, weather forecast data, user behavior patterns, and energy consumption target data, collecting multi-factor, multi-source data.

[0019] Clean the multi-factor, multi-source data to remove noise and reduce its interference with intelligent temperature adaptive regulation.

[0020] Transform multi-factor, multi-source data to remove dimensional differences and form standardized multi-factor, multi-source data.

[0021] Feature extraction is performed on multi-factor, multi-source data. Feature vectors related to intelligent temperature adaptive regulation are extracted from the multi-factor, multi-source data, and the extracted feature vectors are weighted and fused to form multi-factor fused data.

[0022] Preferably, ambient temperature data is collected in real time and transmitted to the cloud platform via a wireless network to perform the following operations:

[0023] Deploy multiple IoT temperature sensors in the area that needs to be monitored, and collect ambient temperature data in real time based on the multiple IoT temperature sensors;

[0024] Among them, the IoT temperature sensor is either a non-contact infrared sensor or a contact thermistor sensor, and the appropriate sensor type is selected according to the specific scenario.

[0025] The IoT temperature sensor transmits the real-time ambient temperature data to the gateway via a wireless network. The gateway collects the ambient temperature data and performs preliminary processing.

[0026] The gateway uploads the pre-processed ambient temperature data to the cloud platform via a wireless network for remote monitoring and storage of the ambient temperature data.

[0027] Preferably, a multi-factor threshold recommendation model is constructed using machine learning algorithms, and the following operations are performed:

[0028] Collect multi-factor historical data and divide the collected multi-factor historical data into training set and test set;

[0029] Based on machine learning technology, a training set is used to train the machine learning model, enabling the machine learning model to learn autonomously from the training set to dynamically adjust or recommend threshold behaviors. Specifically, the threshold is dynamically adjusted or recommended based on historical data, weather forecasts, user behavior patterns, and energy consumption targets, thus determining a multi-factor threshold recommendation model.

[0030] The multi-factor threshold recommendation model is tested based on the test set to evaluate its performance and determine whether it can achieve the expected effect of dynamic adjustment or recommendation threshold, thereby determining the model test evaluation results.

[0031] Based on the model testing and evaluation results, the parameters of the multi-factor threshold recommendation model are adjusted and optimized to determine the optimal multi-factor threshold recommendation model.

[0032] Preferably, the ambient temperature data is analyzed to determine whether the current ambient temperature exceeds the recommended temperature threshold range, and the following operations are performed:

[0033] After receiving ambient temperature data, the cloud platform analyzes the data based on recommended temperature thresholds to determine whether the current ambient temperature exceeds the recommended threshold range, identifies the trend of temperature change, and determines the threshold analysis result.

[0034] When the ambient temperature data is within the temperature threshold range, the threshold analysis result is that the threshold judgment is normal and no temperature adjustment is required.

[0035] When the ambient temperature data is outside the temperature threshold range, the threshold analysis result is that the threshold judgment is abnormal and temperature adjustment is required.

[0036] Based on the degree to which the ambient temperature data exceeds the temperature threshold, it is divided into three levels: slightly exceeding, moderately exceeding, and severely exceeding.

[0037] Preferably, based on the threshold analysis results, different adjustment intensities are adopted, and the following operations are performed:

[0038] The adjustment intensity varies depending on how much the ambient temperature exceeds the temperature threshold. For a slight exceedance, only the equipment power is finely adjusted, the air conditioner fan speed is reduced from high to medium, and the heating temperature is lowered by 1°C. For a moderate exceedance, auxiliary equipment is activated, and the main air conditioner and auxiliary fan circulation are turned on. For a severe exceedance, the core equipment is run at full power, the maximum fan speed of the air conditioner is adjusted, and the windows are closed to reduce heat exchange. The air conditioner, heating system, fan, and window equipment are controlled to work together.

[0039] Preferably, when adjusting the temperature, if there is a conflict in the demand of different areas, resources are dynamically allocated according to preset priorities to intelligently adjust the temperature.

[0040] Preferably, it also includes an evaluation module, used to evaluate the recommendation threshold within a preset time period, determine the evaluation result, and intervene and adjust the recommendation threshold for the next time based on the evaluation result;

[0041] The evaluation module includes:

[0042] The fit evaluation submodule is used for:

[0043] Obtain the threshold temperature for recommending thresholds to customers within a preset time period, as well as the actual ambient temperature at the time of each threshold recommendation;

[0044] The average value of the actual ambient temperature within a preset time period is determined based on the actual ambient temperature.

[0045] The temperature deviation between the actual temperature and the threshold temperature is determined based on the ratio of the average actual ambient temperature within the preset time period to the threshold temperature.

[0046] The degree of fit between the actual temperature mean and the threshold is calculated using a quadratic function based on the deviation between the actual temperature mean and the threshold.

[0047] The subjective approval rating submodule is used for:

[0048] Obtain the number of valid satisfaction ratings for customer recommendations and the total number of recommendation responses; use the ratio of the number of valid satisfaction ratings to the total number of recommendation responses as the percentage of satisfaction ratings.

[0049] Obtain the lowest satisfaction rating, highest satisfaction rating, and satisfaction rating for each key interaction from historical data;

[0050] The key interaction feedback index is determined based on the lowest satisfaction score, the highest satisfaction score, and the satisfaction score for each key interaction.

[0051] The product of the percentage of satisfied responses and the key interaction feedback index is used as the customer's subjective acceptance of the recommendation threshold.

[0052] The evaluation submodule is used to determine the evaluation value of the recommended threshold within a preset time period based on the fit between the actual average temperature and the threshold and the customer's subjective acceptance of the recommended threshold.

[0053]

[0054] Where P represents the evaluation value of the recommended threshold within a preset time period; T actual,i The temperature represents the actual ambient temperature at time i; n represents the length of the preset time period; T target This indicates the recommended threshold temperature for users; S good This indicates the number of valid satisfaction ratings for customer feedback on recommendations; S all S represents the total number of responses to recommendations within a preset time period. key,j S represents the satisfaction score of the j-th key interaction; k represents the total number of key interactions; S min S represents the lowest satisfaction rating in historical data; max This represents the highest satisfaction rating in historical data;

[0055] The intervention adjustment submodule is used to compare the evaluation value with a preset evaluation threshold. If the evaluation value is determined to be less than the preset evaluation threshold, the next recommended threshold is adjusted proportionally based on a preset adjustment coefficient.

[0056] Preferably, it also includes a supplementary training submodule, used to supplement the training of the multi-factor threshold recommendation model to obtain an optimized multi-factor threshold recommendation model;

[0057] The supplementary training submodule includes:

[0058] The data acquisition unit is used to collect all historical threshold recommendation events in IoT temperature regulation scenarios; the historical threshold recommendation events include environmental attributes, device parameters, recommendation information and feedback data;

[0059] The classification unit is used to classify historical threshold recommendation events into proactive intervention recommendation events and system-triggered recommendation events based on the recommendation triggering mechanism.

[0060] An active intervention feature extraction unit is used to extract feature tags from the initiator of the active recommendation and construct a profile of the recommending subject based on the feature tags; wherein, the feature tags include role type, professional background and historical recommendation performance;

[0061] Active intervention verification unit, used for:

[0062] The system analyzes the success rate of adjustments for the same regional type in historical successful recommendation cases to assess the experience reserves of the recommending entity in similar scenarios, constructing an experience dimension. Based on the scenario's demand for professional knowledge, it evaluates whether the recommending entity's knowledge reserves match, constructing a professionalism dimension. The system calculates the error rate of device adjustments after recommendation based on the deviation between the recommendation threshold and the actual temperature achieved after device adjustment, constructing an effectiveness dimension. A three-dimensional verification model of experience, professionalism, and effectiveness is constructed based on these dimensions. A dynamic weighted algorithm is used to generate a comprehensive qualification score for the recommending entity profile. The comprehensive qualification score is compared with a preset comprehensive qualification score threshold, and sample data with comprehensive qualification scores lower than the preset threshold are deleted.

[0063] A system-triggered feature extraction unit is used to extract credit-related data when the system generates recommendations; the credit-related data includes system stability, historical recommendation compliance, and anomaly response capability.

[0064] System-triggered verification unit, used for:

[0065] Credit data is broken down and arranged along a timeline to analyze risk trends;

[0066] Construct a database of negative credit characteristics and perform feature matching on credit events;

[0067] For the matched adverse features, the total risk value of a single recommended event is calculated by combining the impact degree risk value; the total risk value is compared with a preset risk threshold, and recommended events with a total risk value greater than or equal to the preset risk threshold are deleted;

[0068] The sample integration unit is used to integrate the sample data verified by the active intervention-type verification unit with the sample data verified by the system-triggered verification unit to generate a supplementary training sample dataset.

[0069] The supplementary training unit is used to supplement the training of the multi-factor threshold recommendation model based on the supplementary training sample dataset, so as to obtain the optimized multi-factor threshold recommendation model.

[0070] Compared with the prior art, the beneficial effects of the present invention are:

[0071] 1. This invention collects ambient temperature data in real time through multiple deployed IoT temperature sensors and transmits the collected ambient temperature data to a gateway. The gateway collects the ambient temperature data, performs preliminary processing, and then uploads the ambient temperature data to a cloud platform. A multi-factor threshold recommendation model is constructed using machine learning algorithms, and the threshold is dynamically adjusted or recommended by combining historical data, weather forecasts, user behavior patterns, and energy consumption targets, thereby achieving intelligent management of temperature thresholds.

[0072] 2. This invention analyzes ambient temperature data to determine whether the current ambient temperature exceeds the recommended temperature threshold range, identifies the trend of temperature change, determines the threshold analysis result, and takes different adjustment intensities according to the degree to which the ambient temperature exceeds the temperature threshold. Specifically, when the temperature exceeds the threshold slightly, only the equipment power is finely adjusted; when it exceeds it moderately, auxiliary equipment is activated; and when it exceeds it severely, the core equipment runs at full power, controlling the air conditioner, heating, fan, and window equipment to work together. When there are conflicting needs in multiple areas, adjustments are made according to preset priorities. By monitoring the ambient temperature in real time through Internet of Things technology and adjusting the ambient temperature automatically and intelligently based on the dynamically adjusted temperature threshold, the invention achieves comfortable, energy-saving, and efficient management. Attached Figure Description

[0073] Figure 1 This is a block diagram of the intelligent temperature adaptive regulation system based on IoT temperature threshold analysis of the present invention;

[0074] Figure 2 This is a flowchart of the intelligent temperature adaptive adjustment system based on IoT temperature threshold analysis according to the present invention. Detailed Implementation

[0075] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0076] To address the limitations of existing methods that fail to achieve intelligent and precise temperature control through real-time monitoring, data analysis, and intelligent regulation, and which lack the ability to provide comfortable, energy-efficient, and highly effective adaptive management, please refer to [link to relevant documentation]. Figures 1-2 This embodiment provides the following technical solution:

[0077] An intelligent temperature adaptive regulation system based on IoT temperature threshold analysis includes:

[0078] The data sensing and transmission module is used to collect ambient temperature data in real time and transmit the collected ambient temperature data to the cloud platform via a wireless network.

[0079] In this embodiment, ambient temperature data is collected in real time and transmitted to the cloud platform via a wireless network to perform the following operations:

[0080] Deploy multiple IoT temperature sensors in the area that needs to be monitored, and collect ambient temperature data in real time based on the multiple IoT temperature sensors;

[0081] Among them, the IoT temperature sensor is a non-contact infrared sensor or a contact thermistor sensor. The appropriate sensor type is selected according to the specific scenario, and the sampling frequency is also dynamically adjusted according to the scenario, such as 5 minutes / time in normal scenarios and 1 minute / time in rapidly changing scenarios, to ensure data real-time performance and reliability.

[0082] The IoT temperature sensor transmits the real-time ambient temperature data to the gateway via a wireless network. The gateway collects the ambient temperature data and performs preliminary processing.

[0083] The gateway uploads the pre-processed ambient temperature data to the cloud platform via a wireless network for remote monitoring and storage of ambient temperature data, supporting massive data storage and global analysis.

[0084] The threshold dynamic adjustment module is used to dynamically adjust or recommend thresholds based on historical data, weather forecasts, user behavior patterns, and energy consumption targets, enabling intelligent management of thresholds.

[0085] In this embodiment, the threshold is dynamically adjusted or recommended based on historical data, weather forecasts, user behavior patterns, and energy consumption targets, and the following operations are performed:

[0086] Collect and process multi-factor, multi-source data to determine multi-factor fusion data;

[0087] Among them, real-time temperature data, user-preset thresholds, historical operation data, weather forecast data, user behavior patterns and energy consumption target data are monitored, and multi-factor and multi-source data are collected;

[0088] Clean the multi-factor, multi-source data to remove noise and reduce its interference with intelligent temperature adaptive regulation.

[0089] Transform multi-factor, multi-source data to remove dimensional differences and form standardized multi-factor, multi-source data.

[0090] Feature extraction is performed on multi-factor, multi-source data. Feature vectors related to intelligent temperature adaptive regulation are extracted from the multi-factor, multi-source data, and the extracted feature vectors are weighted and fused to form multi-factor fused data.

[0091] A multi-factor threshold recommendation model is constructed using machine learning algorithms;

[0092] This involves collecting multi-factor historical data and dividing the collected multi-factor historical data into training and test sets.

[0093] Based on machine learning technology, a training set is used to train the machine learning model, enabling the machine learning model to learn autonomously from the training set to dynamically adjust or recommend threshold behaviors. Specifically, the threshold is dynamically adjusted or recommended based on historical data, weather forecasts, user behavior patterns, and energy consumption targets, thus determining a multi-factor threshold recommendation model.

[0094] The multi-factor threshold recommendation model is tested based on the test set to evaluate its performance and determine whether it can achieve the expected effect of dynamic adjustment or recommendation threshold, thereby determining the model test evaluation results.

[0095] Based on the model testing and evaluation results, the parameters of the multi-factor threshold recommendation model are adjusted and optimized to determine the optimal multi-factor threshold recommendation model.

[0096] Deploy the optimal multi-factor threshold recommendation model, which involves deploying the multi-factor threshold recommendation model in a real multi-factor threshold recommendation environment;

[0097] Multi-factor fusion data is input into a multi-factor threshold recommendation model, and the multi-factor fusion data is analyzed based on the multi-factor threshold recommendation model. The threshold is dynamically adjusted or recommended based on historical data, weather forecasts, user behavior patterns and energy consumption targets.

[0098] By analyzing users' daily activity time, the system automatically switches the corresponding scene thresholds, analyzes the temperature change patterns of the same period over the past 30 days, predicts temperature fluctuations and adjusts the thresholds in advance, and dynamically corrects the indoor thresholds by combining the outdoor temperature and light intensity for the next 24 hours. When the real-time energy consumption approaches the preset monthly and annual targets, the threshold range is automatically tightened to determine the recommended temperature threshold, thus achieving intelligent management of temperature thresholds.

[0099] The threshold analysis and judgment module is used to analyze ambient temperature data and determine whether the current ambient temperature data exceeds the recommended temperature threshold range, and to determine the threshold analysis and judgment result.

[0100] In this embodiment, the ambient temperature data is analyzed to determine whether the current ambient temperature exceeds the recommended temperature threshold range, and the following operations are performed:

[0101] After receiving ambient temperature data, the cloud platform analyzes the data based on recommended temperature thresholds to determine whether the current ambient temperature exceeds the recommended threshold range, identifies the trend of temperature change, and determines the threshold analysis result.

[0102] When the ambient temperature data is within the temperature threshold range, the threshold analysis result is that the threshold judgment is normal and no temperature adjustment is required.

[0103] When the ambient temperature data is outside the temperature threshold range, the threshold analysis result is that the threshold judgment is abnormal and temperature adjustment is required.

[0104] Based on the degree to which the ambient temperature data exceeds the temperature threshold, it is divided into three levels: slightly exceeding, moderately exceeding, and severely exceeding.

[0105] Specifically, the ambient temperature data is analyzed based on the recommended temperature threshold to determine whether the current ambient temperature exceeds the recommended temperature threshold range. The threshold analysis results are shown in Table 1.

[0106] Table 1: Threshold Analysis and Judgment Results

[0107]

[0108] Therefore, by analyzing ambient temperature data based on recommended temperature thresholds, determining whether the current ambient temperature exceeds the recommended temperature threshold range, and identifying the trend of temperature change, the threshold analysis results can be determined. This facilitates the subsequent adoption of different adjustment intensities based on the threshold analysis results, ensuring intelligent temperature adaptive adjustment.

[0109] The coordination control and adjustment module is used to adopt different adjustment intensities based on the threshold analysis results, so as to realize multi-device collaborative control and conflict coordination, and ensure intelligent temperature adaptive adjustment.

[0110] In this embodiment, different adjustment intensities are applied based on the threshold analysis results, and the following operations are performed:

[0111] The adjustment intensity varies depending on how much the ambient temperature exceeds the temperature threshold. For a slight exceedance, only the equipment power is finely adjusted, the air conditioner fan speed is reduced from high to medium, and the heating temperature is lowered by 1°C. For a moderate exceedance, auxiliary equipment is activated, and the main air conditioner and auxiliary fan circulation are turned on. For a severe exceedance, the core equipment is run at full power, the maximum fan speed of the air conditioner is adjusted, and the windows are closed to reduce heat exchange. The air conditioner, heating system, fan, and window equipment are controlled to work together.

[0112] In this embodiment, when adjusting the temperature, if there is a conflict in the demand from different areas, resources are dynamically allocated according to a preset priority to intelligently adjust the temperature.

[0113] It should be noted that different adjustment intensities are applied based on the degree to which the ambient temperature exceeds the temperature threshold. The intelligent temperature adaptive adjustment is shown in Table 2.

[0114] Table 2: Intelligent Temperature Adaptive Regulation

[0115]

[0116] Therefore, by adjusting the ambient temperature to different degrees as the ambient temperature exceeds the temperature threshold, different adjustment intensities can be adopted, enabling automatic and intelligent adjustment of the ambient temperature and achieving comfortable, energy-saving, and efficient management.

[0117] In summary, by analyzing ambient temperature data, it can determine whether the current ambient temperature exceeds the recommended temperature threshold range, identify the trend of temperature change, determine the threshold analysis results, and take different adjustment intensities according to the different degrees to which the ambient temperature exceeds the temperature threshold. When there are conflicting needs in multiple areas, adjustments are made according to preset priorities. Ambient temperature is monitored in real time through IoT technology, and the ambient temperature is automatically and intelligently adjusted based on dynamically adjusted temperature thresholds to achieve comfortable, energy-saving, and efficient management.

[0118] In this embodiment, it also includes an evaluation module, used to evaluate the recommendation threshold within a preset time period, determine the evaluation result, and intervene and adjust the recommendation threshold for the next time based on the evaluation result;

[0119] The evaluation module includes:

[0120] The fit evaluation submodule is used for:

[0121] Obtain the threshold temperature for recommending thresholds to customers within a preset time period, as well as the actual ambient temperature at the time of each threshold recommendation;

[0122] The average value of the actual ambient temperature within a preset time period is determined based on the actual ambient temperature.

[0123] The temperature deviation between the actual temperature and the threshold temperature is determined based on the ratio of the average actual ambient temperature within the preset time period to the threshold temperature.

[0124] The degree of fit between the actual temperature mean and the threshold is calculated using a quadratic function based on the deviation between the actual temperature mean and the threshold.

[0125] The subjective approval rating submodule is used for:

[0126] Obtain the number of valid satisfaction ratings for customer recommendations and the total number of recommendation responses; use the ratio of the number of valid satisfaction ratings to the total number of recommendation responses as the percentage of satisfaction ratings.

[0127] Obtain the lowest satisfaction rating, highest satisfaction rating, and satisfaction rating for each key interaction from historical data;

[0128] The key interaction feedback index is determined based on the lowest satisfaction score, the highest satisfaction score, and the satisfaction score for each key interaction.

[0129] The product of the percentage of satisfied responses and the key interaction feedback index is used as the customer's subjective acceptance of the recommendation threshold.

[0130] The evaluation submodule is used to determine the evaluation value of the recommended threshold within a preset time period based on the fit between the actual average temperature and the threshold and the customer's subjective acceptance of the recommended threshold.

[0131]

[0132] Where P represents the evaluation value of the recommended threshold within a preset time period; T actual,i The temperature represents the actual ambient temperature at time i; n represents the length of the preset time period; T target This indicates the recommended threshold temperature for users; S good This indicates the number of valid satisfaction ratings for customer feedback on recommendations; S all S represents the total number of responses to recommendations within a preset time period. key,j S represents the satisfaction score of the j-th key interaction; k represents the total number of key interactions; S min S represents the lowest satisfaction rating in historical data; max This represents the highest satisfaction rating in historical data;

[0133] The intervention adjustment submodule is used to compare the evaluation value with a preset evaluation threshold. If the evaluation value is determined to be less than the preset evaluation threshold, the next recommended threshold is adjusted proportionally based on a preset adjustment coefficient.

[0134] In this embodiment, the temperature deviation between the actual temperature and the threshold temperature is determined based on the ratio of the average actual ambient temperature within the preset time period to the threshold temperature, i.e.:

[0135] In this embodiment, the fit between the actual average temperature and the threshold is calculated using a quadratic function based on the deviation between the actual average temperature and the threshold; that is, the quadratic function form.

[0136] In this embodiment, key interactions refer to user behaviors or scenario interactions that have the most direct impact on the effectiveness and reasonableness of the recommendation threshold. For example, in air conditioning temperature recommendations, "the user actually sets the temperature according to the recommended threshold and uses it for more than 2 hours" is a key interaction; while "the user only views the recommended threshold but does not adjust the air conditioner" is a non-key interaction. The core reason for introducing key interactions is to solve the problem that ordinary interactions cannot accurately reflect the true value of the recommendation threshold.

[0137] In this embodiment, the score for key interactions is a quantitative description of the quality of a single key interaction, used to distinguish the degree to which different key interactions support the recommendation threshold. For example, even when a user performs the same key interaction according to the recommendation threshold, there may be differences: Interaction 1: After the user performs the operation according to the recommendation threshold, the feedback is "very suitable, great experience" (can be rated 9 out of 10); Interaction 2: After the user performs the operation according to the recommendation threshold, the feedback is "barely acceptable, but slightly uncomfortable" (can be rated 6 out of 10). Both of these are key interactions, but their quality is different. The role of the score is to reflect this difference numerically. The specific significance of introducing the score is to solve the limitation of simply counting key interactions without distinguishing their quality.

[0138] In this embodiment, the core objective of the formula is to quantify the comprehensive rationality of the recommendation threshold, while also taking into account the adaptability to the customer's subjective environment (temperature dimension) and the user's subjective acceptance (satisfaction dimension).

[0139] The working principle and beneficial effects of the above technical solution are as follows: The fit evaluation submodule assesses the fit between the actual ambient temperature and the recommended threshold temperature, taking into account the impact of environmental factors on the recommended threshold. If the actual temperature deviates significantly from the recommended threshold, it indicates that the recommended threshold may be unsuitable and can be adjusted, thus improving the matching degree between the recommended threshold and the actual environment, and consequently improving the accuracy of the recommendation. The subjective acceptance evaluation submodule comprehensively evaluates the customer's subjective acceptance of the recommended threshold based on customer satisfaction feedback and satisfaction scores of key interactions. The evaluation submodule combines fit and subjective acceptance to obtain an evaluation value, and the intervention and adjustment submodule adjusts the next recommended threshold based on the evaluation value. This approach better meets customer needs and improves customer satisfaction because it considers not only objective environmental factors but also the customer's subjective feelings. The entire solution forms a closed-loop evaluation and adjustment mechanism. By evaluating the recommended threshold within a preset time period, problems in the recommendation strategy can be identified in a timely manner, and then the next recommended threshold can be adjusted based on the evaluation results. This allows for continuous optimization of the recommendation strategy, making the recommendation system more intelligent and efficient.

[0140] In this embodiment, a supplementary training submodule is also included, which is used to supplement the training of the multi-factor threshold recommendation model to obtain an optimized multi-factor threshold recommendation model.

[0141] The supplementary training submodule includes:

[0142] The data acquisition unit is used to collect all historical threshold recommendation events in IoT temperature regulation scenarios; the historical threshold recommendation events include environmental attributes, device parameters, recommendation information and feedback data;

[0143] The classification unit is used to classify historical threshold recommendation events into proactive intervention recommendation events and system-triggered recommendation events based on the recommendation triggering mechanism.

[0144] An active intervention feature extraction unit is used to extract feature tags from the initiator of the active recommendation and construct a profile of the recommending subject based on the feature tags; wherein, the feature tags include role type, professional background and historical recommendation performance;

[0145] Active intervention verification unit, used for:

[0146] The system analyzes the success rate of adjustments for the same regional type in historical successful recommendation cases to assess the experience reserves of the recommending entity in similar scenarios, constructing an experience dimension. Based on the scenario's demand for professional knowledge, it evaluates whether the recommending entity's knowledge reserves match, constructing a professionalism dimension. The system calculates the error rate of device adjustments after recommendation based on the deviation between the recommendation threshold and the actual temperature achieved after device adjustment, constructing an effectiveness dimension. A three-dimensional verification model of experience, professionalism, and effectiveness is constructed based on these dimensions. A dynamic weighted algorithm is used to generate a comprehensive qualification score for the recommending entity profile. The comprehensive qualification score is compared with a preset comprehensive qualification score threshold, and sample data with comprehensive qualification scores lower than the preset threshold are deleted.

[0147] A system-triggered feature extraction unit is used to extract credit-related data when the system generates recommendations; the credit-related data includes system stability, historical recommendation compliance, and anomaly response capability.

[0148] System-triggered verification unit, used for:

[0149] Credit data is broken down and arranged along a timeline to analyze risk trends;

[0150] Construct a database of negative credit characteristics and perform feature matching on credit events;

[0151] For the matched adverse features, the total risk value of a single recommended event is calculated by combining the impact degree risk value; the total risk value is compared with a preset risk threshold, and recommended events with a total risk value greater than or equal to the preset risk threshold are deleted;

[0152] The sample integration unit is used to integrate the sample data verified by the active intervention-type verification unit with the sample data verified by the system-triggered verification unit to generate a supplementary training sample dataset.

[0153] The supplementary training unit is used to supplement the training of the multi-factor threshold recommendation model based on the supplementary training sample dataset, so as to obtain the optimized multi-factor threshold recommendation model.

[0154] In this embodiment, environmental attributes include area type, space size, and real-time temperature and humidity; equipment parameters include equipment type, adjustment range, and energy consumption index; recommendation information includes recommendation threshold, recommendation time, and execution result; and feedback data includes equipment adjustment error, user satisfaction, and energy consumption changes.

[0155] In this embodiment, proactive intervention recommendations are initiated manually, such as users customizing thresholds based on specific needs (e.g., constant temperature in a baby room) or engineers manually adjusting parameters for industrial scenarios; system-triggered recommendations are automatically generated by the system, such as threshold updates based on periodic detection or adaptive recommendations when environmental parameters change abruptly.

[0156] In this embodiment, the role types include, but are not limited to, ordinary users, HVAC engineers, and system administrators; the professional backgrounds include, but are not limited to, industrial temperature control experience and proficiency in operating home smart devices; and the historical recommendation performance includes, but is not limited to, the number of successful cases, the adjustment error rate, and the scene matching degree.

[0157] In this embodiment, system stability includes whether there are faults or data transmission delays during the recommended time period; historical recommendation compliance includes whether the recommended threshold is within the safe operating range of the equipment; and abnormal response capability includes the speed of recommendation adjustment to sudden temperature fluctuations.

[0158] The working principle and beneficial effects of the above technical solution are as follows: Historical threshold recommendation events are filtered through proactive intervention and system-triggered verification units, removing data that does not meet the requirements; this ensures the quality of the sample dataset used for supplementary training, avoiding interference from low-quality data and thus improving model accuracy; the supplementary training unit trains the multi-factor threshold recommendation model based on a carefully selected and integrated supplementary training sample dataset, enabling the model to better learn effective information in different scenarios. Both proactive intervention and system-triggered effective data are integrated and utilized, helping to optimize the model's recommendation performance in IoT temperature regulation scenarios and improve the accuracy of recommendation thresholds; the experience-expertise-effect three-dimensional verification model in proactive intervention verification and the credit correlation analysis in system-triggered verification evaluate the data from multiple perspectives. This comprehensive evaluation and screening process enables the model to learn based on more reliable data during training, thereby improving the model's reliability in practical applications and reducing the possibility of erroneous recommendations. Since the sample data covers a variety of situations, including both proactive intervention and system-triggered scenarios, and has undergone rigorous screening and processing, the optimized multi-factor threshold recommendation model can better adapt to various complex situations in IoT temperature regulation scenarios. Whether the recommendation needs are driven by proactive human intervention or automatically triggered by the system, the model can make more reasonable responses.

[0159] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0160] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An intelligent temperature adaptive control system based on IoT temperature threshold analysis, characterized in that, include: The data sensing and transmission module is used to collect ambient temperature data in real time and transmit the collected ambient temperature data to the cloud platform via a wireless network. The threshold dynamic adjustment module is used to dynamically adjust or recommend thresholds based on historical data, weather forecasts, user behavior patterns, and energy consumption targets, for intelligent management of thresholds. The threshold analysis and judgment module is used to analyze the ambient temperature data and determine whether the current ambient temperature data exceeds the recommended temperature threshold range, and to determine the threshold analysis and judgment result. The coordination control and adjustment module is used to adopt different adjustment intensities based on the threshold analysis results, so as to realize multi-device collaborative control and conflict coordination, and ensure intelligent temperature adaptive adjustment.

2. The intelligent temperature adaptive regulation system based on IoT temperature threshold analysis according to claim 1, characterized in that, Based on historical data, weather forecasts, user behavior patterns, and energy consumption targets, dynamically adjust or recommend thresholds and perform the following operations: Collect and process multi-factor, multi-source data to determine multi-factor fusion data; A multi-factor threshold recommendation model is constructed using machine learning algorithms, and the multi-factor threshold recommendation model is deployed in a real multi-factor threshold recommendation environment. Multi-factor fusion data is input into a multi-factor threshold recommendation model, and the multi-factor fusion data is analyzed based on the multi-factor threshold recommendation model. The threshold is dynamically adjusted or recommended based on historical data, weather forecasts, user behavior patterns and energy consumption targets. By analyzing users' daily activity time, the system automatically switches the corresponding scene thresholds, analyzes the temperature change patterns of the same period over the past 30 days, predicts temperature fluctuations and adjusts the thresholds in advance, and dynamically corrects the indoor thresholds by combining the outdoor temperature and light intensity for the next 24 hours. When the real-time energy consumption is close to the preset monthly and annual targets, the system automatically tightens the threshold range and determines the recommended temperature threshold.

3. The intelligent temperature adaptive regulation system based on IoT temperature threshold analysis according to claim 2, characterized in that, Collect and process multi-factor, multi-source data to determine the multi-factor fusion data, and perform the following operations: It monitors real-time temperature data, user-preset thresholds, historical operating data, weather forecast data, user behavior patterns, and energy consumption target data, collecting multi-factor, multi-source data. Clean the multi-factor, multi-source data to remove noise and reduce its interference with intelligent temperature adaptive regulation. Transform multi-factor, multi-source data to remove dimensional differences and form standardized multi-factor, multi-source data. Feature extraction is performed on multi-factor, multi-source data. Feature vectors related to intelligent temperature adaptive regulation are extracted from the multi-factor, multi-source data, and the extracted feature vectors are weighted and fused to form multi-factor fused data.

4. The intelligent temperature adaptive regulation system based on IoT temperature threshold analysis according to claim 3, characterized in that, Real-time ambient temperature data is collected and transmitted to the cloud platform via wireless network. The following operations are then performed: Deploy multiple IoT temperature sensors in the area that needs to be monitored, and collect ambient temperature data in real time based on the multiple IoT temperature sensors; Among them, the IoT temperature sensor is either a non-contact infrared sensor or a contact thermistor sensor, and the appropriate sensor type is selected according to the specific scenario. The IoT temperature sensor transmits the real-time ambient temperature data to the gateway via a wireless network. The gateway collects the ambient temperature data and performs preliminary processing. The gateway uploads the pre-processed ambient temperature data to the cloud platform via a wireless network for remote monitoring and storage of the ambient temperature data.

5. The intelligent temperature adaptive regulation system based on IoT temperature threshold analysis according to claim 4, characterized in that, A multi-factor threshold recommendation model is constructed using machine learning algorithms, and the following operations are performed: Collect multi-factor historical data and divide the collected multi-factor historical data into training set and test set; Based on machine learning technology, a training set is used to train the machine learning model, enabling the machine learning model to learn autonomously from the training set to dynamically adjust or recommend threshold behaviors. Specifically, the threshold is dynamically adjusted or recommended based on historical data, weather forecasts, user behavior patterns, and energy consumption targets, thus determining a multi-factor threshold recommendation model. The multi-factor threshold recommendation model is tested based on the test set to evaluate its performance and determine whether it can achieve the expected effect of dynamic adjustment or recommendation threshold, thereby determining the model test evaluation results. Based on the model testing and evaluation results, the parameters of the multi-factor threshold recommendation model are adjusted and optimized to determine the optimal multi-factor threshold recommendation model.

6. The intelligent temperature adaptive regulation system based on IoT temperature threshold analysis according to claim 5, characterized in that, Analyze the ambient temperature data and determine whether the current ambient temperature exceeds the recommended temperature threshold range. Perform the following operations: After receiving ambient temperature data, the cloud platform analyzes the data based on recommended temperature thresholds to determine whether the current ambient temperature exceeds the recommended threshold range, identifies the trend of temperature change, and determines the threshold analysis result. When the ambient temperature data is within the temperature threshold range, the threshold analysis result is that the threshold judgment is normal and no temperature adjustment is required. When the ambient temperature data is outside the temperature threshold range, the threshold analysis result is that the threshold judgment is abnormal and temperature adjustment is required. Based on the degree to which the ambient temperature data exceeds the temperature threshold, it is divided into three levels: slightly exceeding, moderately exceeding, and severely exceeding.

7. The intelligent temperature adaptive regulation system based on IoT temperature threshold analysis according to claim 6, characterized in that, Based on the threshold analysis results, different adjustment intensities are applied, and the following operations are performed: The adjustment intensity varies depending on how much the ambient temperature exceeds the temperature threshold. For a slight exceedance, only the equipment power is finely adjusted, the air conditioner fan speed is reduced from high to medium, and the heating temperature is lowered by 1°C. For a moderate exceedance, auxiliary equipment is activated, and the main air conditioner and auxiliary fan circulation are turned on. For a severe exceedance, the core equipment is run at full power, the maximum fan speed of the air conditioner is adjusted, and the windows are closed to reduce heat exchange. The air conditioner, heating system, fan, and window equipment are controlled to work together.

8. The intelligent temperature adaptive regulation system based on IoT temperature threshold analysis according to claim 7, characterized in that, When adjusting the temperature, if there is a conflict in the demand from different areas, resources are dynamically allocated to intelligently adjust the temperature according to the preset priority.

9. The intelligent temperature adaptive regulation system based on IoT temperature threshold analysis according to claim 1, characterized in that, It also includes an evaluation module, which is used to evaluate the recommendation threshold within a preset time period and determine the evaluation result; The next recommendation threshold will be adjusted based on the evaluation results. The evaluation module includes: The fit evaluation submodule is used for: Obtain the threshold temperature for recommending thresholds to customers within a preset time period, as well as the actual ambient temperature at the time of each threshold recommendation; The average value of the actual ambient temperature within a preset time period is determined based on the actual ambient temperature. The temperature deviation between the actual temperature and the threshold temperature is determined based on the ratio of the average actual ambient temperature within the preset time period to the threshold temperature. The degree of fit between the actual temperature mean and the threshold is calculated using a quadratic function based on the deviation between the actual temperature mean and the threshold. The subjective approval rating submodule is used for: Obtain the number of valid satisfaction ratings for customer recommendations and the total number of recommendation responses; use the ratio of the number of valid satisfaction ratings to the total number of recommendation responses as the percentage of satisfaction ratings. Obtain the lowest satisfaction rating, highest satisfaction rating, and satisfaction rating for each key interaction from historical data; The key interaction feedback index is determined based on the lowest satisfaction score, the highest satisfaction score, and the satisfaction score for each key interaction. The product of the percentage of satisfied responses and the key interaction feedback index is used as the customer's subjective acceptance of the recommendation threshold. The evaluation submodule is used to determine the evaluation value of the recommended threshold within a preset time period based on the fit between the actual average temperature and the threshold and the customer's subjective acceptance of the recommended threshold. Where P represents the evaluation value of the recommended threshold within a preset time period; T actual,i The temperature represents the actual ambient temperature at time i; n represents the length of the preset time period; T target This indicates the recommended threshold temperature for users; S good This indicates the number of valid satisfaction ratings for customer feedback on recommendations; S all S represents the total number of responses to recommendations within a preset time period. key,j S represents the satisfaction score of the j-th key interaction; k represents the total number of key interactions; S min S represents the lowest satisfaction rating in historical data; max This represents the highest satisfaction rating in historical data; The intervention adjustment submodule is used to compare the evaluation value with a preset evaluation threshold. If the evaluation value is determined to be less than the preset evaluation threshold, the next recommended threshold is adjusted proportionally based on a preset adjustment coefficient.

10. The intelligent temperature adaptive regulation system based on IoT temperature threshold analysis according to claim 2, characterized in that, It also includes a supplementary training submodule, which is used to supplement the training of the multi-factor threshold recommendation model to obtain an optimized multi-factor threshold recommendation model; The supplementary training submodule includes: The data acquisition unit is used to collect all historical threshold recommendation events in IoT temperature regulation scenarios; the historical threshold recommendation events include environmental attributes, device parameters, recommendation information and feedback data; The classification unit is used to classify historical threshold recommendation events into proactive intervention recommendation events and system-triggered recommendation events based on the recommendation triggering mechanism. An active intervention feature extraction unit is used to extract feature tags from the initiator of the active recommendation and construct a profile of the recommending subject based on the feature tags; wherein, the feature tags include role type, professional background and historical recommendation performance; Active intervention verification unit, used for: The system analyzes the success rate of adjustments for the same regional type in historical successful recommendation cases to assess the experience reserves of the recommending entity in similar scenarios, constructing an experience dimension. Based on the scenario's demand for professional knowledge, it evaluates whether the recommending entity's knowledge reserves match, constructing a professionalism dimension. The system calculates the error rate of device adjustments after recommendation based on the deviation between the recommendation threshold and the actual temperature achieved after device adjustment, constructing an effectiveness dimension. A three-dimensional verification model of experience, professionalism, and effectiveness is constructed based on these dimensions. A dynamic weighted algorithm is used to generate a comprehensive qualification score for the recommending entity profile. The comprehensive qualification score is compared with a preset comprehensive qualification score threshold, and sample data with comprehensive qualification scores lower than the preset threshold are deleted. A system-triggered feature extraction unit is used to extract credit-related data when the system generates recommendations; the credit-related data includes system stability, historical recommendation compliance, and anomaly response capability. System-triggered verification unit, used for: Credit data is broken down and arranged along a timeline to analyze risk trends; Construct a database of negative credit characteristics and perform feature matching on credit events; For the matched adverse features, the total risk value of a single recommended event is calculated by combining the impact degree risk value; the total risk value is compared with a preset risk threshold, and recommended events with a total risk value greater than or equal to the preset risk threshold are deleted; The sample integration unit is used to integrate the sample data verified by the active intervention-type verification unit with the sample data verified by the system-triggered verification unit to generate a supplementary training sample dataset. The supplementary training unit is used to supplement the training of the multi-factor threshold recommendation model based on the supplementary training sample dataset, so as to obtain the optimized multi-factor threshold recommendation model.

Images