Intelligent door and window control method and system

Abstract

The invention provides an intelligent door and window control method and system, and the method comprises the steps: S1, obtaining multi-source environment data; s2, inputting the multi-source environment data into a fuzzification module, and converting the multi-source environment data into a fuzzy variable through a preset ambiguity function; s3, dynamically generating an environment parameter weight vector according to the real-time scene parameters; s4, inputting the fuzzy variable and the environmental parameter weight vector into a fuzzy inference engine, executing logical operation of a fuzzy rule base, and outputting comprehensive evaluation values of window opening suitability, window closing urgency and shading demand; and S5, generating a final control instruction through a defuzzification strategy, and executing the final control instruction. According to the method, the problem of frequent invalid operation caused by fluctuation of the environmental parameters near the boundary value is solved, the environmental parameter weight vector is dynamically generated based on the time parameters, the user preset mode and the meteorological data confidence factor, adaptation to seasons, time periods, user states and other scenes is achieved, and decision logic can be flexibly adjusted according to actual requirements; and a contradictory decision under a traditional fixed rule is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of control systems, and in particular to an intelligent door and window control method and system. Background Art

[0002] In the field of intelligent door and window control, the system usually relies on multi-source environmental sensors and meteorological data interfaces to achieve automated decision-making. The environmental parameters involved include temperature and humidity, light intensity, The decision-making goal is to provide users with a comfortable and healthy indoor environment by regulating the switch status of doors and windows and shading devices.

[0003] However, in practical applications, the conflicts and ambiguities of multi-source environmental data often lead to decision-making dilemmas: on the one hand, data conflicts are common, such as the weather API predicting rainfall but the local wind and rain sensor does not detect precipitation, or indoor When the concentration exceeds the standard, windows need to be opened for ventilation, but the outdoor PM2.5 concentration is high; on the other hand, the data ambiguity problem is prominent, such as the temperature fluctuates at the edge of the comfort zone and the light intensity oscillates around the shading threshold.

[0004] Existing technologies mostly use fixed thresholds or simple priority rules to make decisions, which makes it difficult to effectively deal with the above scenarios: for conflicting data, fixed rules are prone to lead to contradictory decisions (such as triggering window opening and closing conditions at the same time); for fuzzy boundary data, hard threshold settings will lead to frequent invalid operations (such as repeatedly opening and closing windows when the temperature fluctuates slightly), which not only misses the optimal control opportunity, but also reduces user comfort and system trust.

[0005] Furthermore, existing solutions lack the ability to dynamically adapt to user context. They are unable to adjust decision logic based on time, season, user activity status (e.g., sleeping, away from home, at home), and mode preferences. This leads to decisions that deviate from human intuition and actual needs. Therefore, an intelligent decision-making mechanism that can handle data conflicts and ambiguity and adapt to dynamic contexts is urgently needed to improve system stability. Summary of the Invention

[0006] Based on the technical problems existing in the above background technology, the present invention proposes an intelligent door and window control method and system, and the technical solutions adopted are as follows:

[0007] A smart door and window control method, the method comprising:

[0008] S1: Acquire multi-source environmental data;

[0009] S2: Input the multi-source environmental data into the fuzzification module and convert it into fuzzy variables through the preset fuzziness function;

[0010] S3: Dynamically generate environmental parameter weight vectors based on real-time scenario parameters;

[0011] S4: Input the fuzzy variables and environmental parameter weight vectors into the fuzzy inference engine, execute the logical operation of the fuzzy rule base, and output the comprehensive evaluation values ​​of window opening suitability, window closing urgency, and shading demand;

[0012] S5: Generate the final control instructions through the defuzzification strategy and execute them.

[0013] Preferably, the multi-source environmental data of S1 includes outdoor temperature and humidity, light intensity, PM2.5 concentration and weather forecast data.

[0014] Preferably, the fuzziness function adopts a trapezoidal function, and the multi-source environmental data is defined to include multiple gradient levels, wherein:

[0015] The fuzziness functions of temperature, humidity, and light dynamically adjust boundary parameters according to the season;

[0016] The fuzziness function of PM2.5 concentration is associated with the health standard threshold.

[0017] Preferably, the weather forecast data is processed in the following manner:

[0018] Generate confidence factors based on the update timeliness of meteorological data and the granularity of meteorological data forecast;

[0019] Convert the predicted rainfall probability and wind speed level into fuzzy variables:

[0020] Rainfall probability gradient: the first gradient is no rain, the second gradient is medium probability rain, and the third gradient is high probability rain;

[0021] Wind speed level gradient: first level is light wind, second level is moderate wind, and third level is strong wind.

[0022] Preferably, the real-time scenario parameters of S3 include:

[0023] Acquire real-time scenario parameters, including:

[0024] Time parameters and user preset modes, wherein the user preset modes include: sleep mode and away mode.

[0025] Preferably, the environmental parameter weight vector is obtained by:

[0026] The preset multi-source environmental data includes weights of outdoor temperature and humidity, light intensity, PM2.5 concentration, and weather forecast data, and the sum of the weights of the multi-source environmental data is 1;

[0027] generating a first adjustment coefficient vector based on the time parameter, for adjusting the weights of outdoor temperature and humidity and light intensity;

[0028] Generate a second adjustment coefficient vector based on a user preset mode, for adjusting the weights of outdoor temperature and humidity, PM2.5 concentration, and weather forecast data;

[0029] A third adjustment coefficient vector is generated based on the confidence factor, and is used to adjust the weight of the weather forecast data.

[0030] Preferably, the first adjustment coefficient vector includes:

[0031] The first sub-coefficient is used to adjust the basic weight of outdoor temperature and humidity;

[0032] The second sub-coefficient is used to adjust the basic weight of light intensity;

[0033] The second adjustment coefficient vector includes:

[0034] The third sub-coefficient is used to adjust the basic weight of outdoor temperature and humidity;

[0035] The fourth sub-coefficient is used to adjust the basic weight of PM2.5 concentration;

[0036] The fifth sub-coefficient is used to adjust the basic weight of weather forecast data.

[0037] Preferably, the S4 specifically includes:

[0038] The fuzzy variables of each environmental parameter and the corresponding environmental parameter weight vector are fused to generate an intermediate fuzzy state;

[0039] Based on the inference rules in the intermediate fuzzy state matching rule library, the corresponding conclusion fuzzy subset is triggered, and the triggered conclusion fuzzy subset is aggregated. The outputs of multiple rules are integrated through weighted summation or the larger or smaller method to generate comprehensive fuzzy evaluation values ​​of window opening suitability, window closing urgency, and shading demand.

[0040] Preferably, the rule base further includes:

[0041] When a rule matching conflict occurs, the conflict resolution rule group is called, the weight ratio of the conflicting rules is adjusted according to the current situation parameters, and instructions that meet the situation priority are generated first.

[0042] An intelligent door and window control system, the system comprising:

[0043] Multi-source environmental perception system: obtains multi-source environmental data;

[0044] Fuzzy variable conversion system: multi-source environmental data is input into the fuzzification module and converted into fuzzy variables through the preset fuzziness function;

[0045] Dynamic weight adaptation system: dynamically generates environmental parameter weight vectors based on real-time scenario parameters;

[0046] Fuzzy reasoning decision system: Fuzzy variables and environmental parameter weight vectors are input into the fuzzy reasoning engine, logical operations of the fuzzy rule base are executed, and comprehensive evaluation values ​​of window opening suitability, window closing urgency, and shading demand are output;

[0047] Control instruction execution system: Generates the final control instruction through defuzzification strategy and executes it.

[0048] Beneficial effects of the present invention: The present invention solves the problem of frequent invalid operations caused by fluctuations in environmental parameters near boundary values, and improves the smoothness and stability of decision-making; by dynamically generating environmental parameter weight vectors based on time parameters, user preset modes and meteorological data confidence factors, it achieves adaptation to scenarios such as seasons, time periods, and user status, so that the decision logic can be flexibly adjusted according to actual needs; it avoids contradictory decisions under traditional fixed rules, and improves the rationality of decision-making in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 This is an intelligent door and window control method described in the present invention. DETAILED DESCRIPTION

[0050] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0051] One embodiment of the present invention provides a smart door and window control method, the method comprising:

[0052] S1: Acquire multi-source environmental data;

[0053] S2: Input the multi-source environmental data into the fuzzification module and convert it into fuzzy variables through the preset fuzziness function;

[0054] S3: Dynamically generate environmental parameter weight vectors based on real-time scenario parameters;

[0055] S4: Input the fuzzy variables and environmental parameter weight vectors into the fuzzy inference engine, execute the logical operation of the fuzzy rule base, and output the comprehensive evaluation values ​​of window opening suitability, window closing urgency, and shading demand;

[0056] S5: Generate the final control instructions through the defuzzification strategy and execute them.

[0057] The working principle and effect of the above technical solution are as follows: the system first obtains multi-source environmental data such as outdoor temperature and humidity, light intensity, PM2.5 concentration, and weather forecast through sensors and meteorological APIs. This data is then input into the fuzzification module and converted into fuzzy variables that conform to natural language descriptions using a preset trapezoidal fuzziness function. This replaces the hard threshold to avoid decision-making oscillations caused by boundary fluctuations. At the same time, the system dynamically generates environmental parameter weight vectors based on real-time scenario parameters. The time parameter adjusts the weights of temperature, humidity, and light through a periodic function and a seasonal amplifier. The user mode adjusts the weights of temperature, humidity, PM2.5, and meteorological data through a mode-specific correction factor. The meteorological data confidence factor further corrects the weights. Afterwards, the fuzzy inference engine inputs the above fuzzy variables and dynamic weight vectors into a preset rule library. By matching multiple sets of condition-conclusion inference rules, logical operations are performed to output comprehensive fuzzy evaluation values ​​for window opening suitability, window closing urgency, and shading demand. Finally, the comprehensive evaluation values ​​are converted into specific control instructions through a defuzzification strategy and executed, realizing the processing of data conflicts and fuzzy boundaries.

[0058] It solves the problem of frequent invalid operations caused by fluctuations in environmental parameters near boundary values, and improves the smoothness and stability of decision-making; by dynamically generating environmental parameter weight vectors based on time parameters, user preset modes and meteorological data confidence factors, it achieves adaptation to scenarios such as seasons, time periods, and user status, so that the decision logic can be flexibly adjusted according to actual needs; it avoids contradictory decisions under traditional fixed rules, and improves the rationality of decisions in complex environments.

[0059] In one embodiment of the present invention, the fuzziness function adopts a trapezoidal function, and defines multi-source environmental data including multiple gradient levels, wherein:

[0060] The fuzziness functions of temperature, humidity, and light dynamically adjust boundary parameters according to the season;

[0061] The fuzziness function of PM2.5 concentration is associated with the health standard threshold.

[0062] The working principle and effect of the above technical solution are as follows: the fuzziness function adopts a trapezoidal function to define three gradient levels for multi-source environmental data. The trapezoidal fuzziness function of temperature sets the comfort zone boundary parameters to [23, 25, 29]°C in summer (corresponding to the gradient of "cool, comfortable, and hot") and adjusts it to [17, 20, 23]°C in winter. For example, when the summer temperature is 28°C, the fuzziness of "comfortable" is 0.25 and the fuzziness of "hot" is 0.75. The trapezoidal function of humidity sets the comfort zone boundaries to [45%, 60%, 75%] in summer (corresponding to "dry, comfortable, and humid") and to [35%, 50%, 65%] in winter. For example, when the humidity is 70% in summer, the fuzziness of "comfortable" is 0.33 and the fuzziness of "humid" is 0.67. The trapezoidal function of light has a strong light threshold of [3000, 10000, 50000] in summer. lux (corresponding to "weak light, moderate light, strong light"), with the range being [2000, 8000, 40000] lux in winter. For example, at 35000 lux at noon in summer, the ambiguity of "strong light" is 0.83. The trapezoidal function of PM2.5 concentration is associated with the health standard threshold, with the boundaries of "good" being [0, 35, 75] μg / m³, "light pollution" being [75, 115, 150] μg / m³, and "moderate pollution" being [150, 250, 350] μg / m³. For example, at a concentration of 90 μg / m³, the ambiguity of "light pollution" is 0.38, and the ambiguity of "moderate pollution" is 0.12.

[0063] The fuzzy function adopts a trapezoidal function with three gradient levels, which not only reduces the computational complexity by simplifying the number of gradients and reduces the computing overhead of the embedded system, but also retains sufficient environmental parameter differentiation and avoids rule redundancy caused by over-segmentation; dynamically adjusts the boundary parameters of temperature, humidity, and light by season, so that the fuzzy division is more in line with the environmental characteristics of different seasons, and improves the adaptability of decisions to actual working conditions; the gradient of PM2.5 concentration is directly related to the health standard threshold, ensuring that the pollution level division meets health protection needs and enhancing the practicality and reliability of decision-making; the three-gradient trapezoidal function can naturally handle the ambiguity of parameters near the boundary, avoid the frequent invalid operations caused by traditional hard thresholds, and improve the stability of system control.

[0064] In one embodiment of the present invention, the weather forecast data is processed in the following manner:

[0065] Generate confidence factors based on the update timeliness of meteorological data and the granularity of meteorological data forecast;

[0066] Convert the predicted rainfall probability and wind speed level into fuzzy variables:

[0067] Rainfall probability gradient: the first gradient is no rain, the second gradient is medium probability rain, and the third gradient is high probability rain;

[0068] Wind speed level gradient: first level is light wind, second level is moderate wind, and third level is strong wind.

[0069] The working principle and effect of the above technical solution are as follows: first, a confidence factor is generated through the update timeliness of meteorological data and the forecast granularity, where the update timeliness is quantified by the "interval from the current time" (the update timeliness weight within 1 hour is 1.0, 2-4 hours is 0.8, 4-6 hours is 0.5, and more than 6 hours is 0.3), and the forecast granularity is quantified by the "time resolution" (the hourly forecast granularity weight is 1.0, 3 hours is 0.7, and daily is 0.4). The confidence factor is the product of the two (if a certain forecast data was updated 2 hours ago and the granularity is 3 hours, then the confidence factor = 0.8×0.7=0.56); then the predicted rainfall probability is converted into a three-gradient fuzzy variable, the first gradient corresponds to no rain 0-20% (trapezoidal function boundary [0,5,15,20]%, and the rainfall probability is 10%). The first gradient corresponds to 0-3 levels (boundary [0,1,2,3], e.g., at level 2, the ambiguity of light wind is 0.5 and the ambiguity of medium wind is 0.5); the second gradient corresponds to 3-6 levels (boundary [2,3,5,6], e.g., at level 4, the ambiguity of medium wind is 0.67 and the ambiguity of light wind is 0.33); the third gradient corresponds to 60-100% (boundary [50,60,90,100]), e.g., at level 80, the ambiguity of medium rain is 0.33 and the ambiguity of high rain is 0.67); the wind speed level is converted into a fuzzy variable with three gradients. The first gradient corresponds to 0-3 levels (boundary [0,1,2,3], e.g., at level 2, the ambiguity of light wind is 0.5 and the ambiguity of medium wind is 0.5); the second gradient corresponds to 3-6 levels (boundary [2,3,5,6], e.g., at level 4, the ambiguity of medium wind is 0.67 and the ambiguity of light wind is 0.33); the third gradient corresponds to 6-10 levels (boundary [50,60,90,100], e.g., at level 4, the ambiguity of strong wind is 0.67 and the ambiguity of strong wind is 0.33). Level (boundary [5, 6, 8, 10] level, strong wind ambiguity is 0.33 when the wind level is 7, and moderate wind ambiguity is 0.67); finally, these fuzzy variables are combined with the confidence factor (for example, the fuzzy variable value of high probability rain 0.67 is multiplied by the confidence factor 0.56 to obtain the actual influence weight of this parameter on the window closing decision);

[0070] The weather forecast data processing method improves decision reliability by combining confidence factors with fuzzy gradients: the confidence factor dynamically identifies data reliability based on update timeliness and forecast granularity, avoiding misoperation caused by reliance on unreliable data; the rainfall probability and wind speed level are converted into three gradient fuzzy variables, and a trapezoidal function is used to achieve smooth boundary transition to avoid decision mutations caused by hard thresholds.

[0071] In one embodiment of the present invention, the environmental parameter weight vector is obtained by:

[0072] The preset multi-source environmental data includes the weights of outdoor temperature and humidity, light intensity, PM2.5 concentration and weather forecast data, and the sum of the weights of each multi-source environmental data is 1; wherein, the preset multi-source environmental data includes ,in, represents the outdoor temperature and humidity weight, represents the light intensity weight, represents the PM2.5 concentration weight, Indicates the weight of weather forecast data;

[0073] A first adjustment coefficient vector is generated based on the time parameter to adjust the weights of the outdoor temperature and humidity and the light intensity; the first adjustment coefficient vector is [f(t)×g(t), h(t), 0, 0], where f(t) represents the outdoor temperature and humidity time factor, and g(t) represents the seasonal factor, and when the season is spring and summer, g(t) = 1.2, and when the season is autumn and winter, g(t) = 0.8; h(t) represents the light intensity time factor, and, ),in Indicates sunrise time, Indicates sunset time;

[0074] A second adjustment coefficient vector is generated based on the user preset mode to adjust the weights of outdoor temperature and humidity, PM2.5 concentration and weather forecast data; the second adjustment coefficient vector is: [ , where m represents the user preset mode, and m is divided into away mode and sleep mode; represents the preset mode factor of outdoor temperature and humidity, and, represents the temperature weight correction factor, , represents the preset mode factor of PM2.5 concentration, and ,in, Expressing health concerns, represents the preset mode factor of the weather forecast data, and, ,in, Expressing safety concerns, .

[0075] A third adjustment coefficient vector is generated based on the confidence factor, and is used to adjust the weight of the weather forecast data.

[0076] The working principle and effect of the above technical solution are as follows: In the natural environment, the temperature shows a "periodic sinusoidal fluctuation" with the change of day and night (such as low temperature in the early morning and high temperature in the afternoon), and Captures this pattern:

[0077] The period is 24 hours (consistent with the Earth's rotation period), with a phase offset of 14 o'clock (14 o'clock is the daily high temperature peak, which is consistent with the actual temperature curve in most regions); the function value range is [0,1] (the cosine function takes the value [-1,1], and after superimposing 0.5, it is mapped to 0-1). It quantifies the "dynamic change of temperature influence weight with day and night" and replaces the artificial "day / night" fixed value with a physical periodic function, avoiding the "sudden change" problem of traditional piecewise functions.

[0078] At sunrise and sunset, the light intensity changes smoothly (rather than suddenly bright / dark). The "S-shaped curve" of the sigmoid function matches this physical process: when it is sunrise, the function value rises smoothly from 0 to 1 (simulating the gradual increase in light during sunrise). When , the function value rises from 0 to 1, and the "light effective time window" is obtained by subtraction. This function replaces the traditional step-by-step judgment of "turning on immediately after sunrise and turning off immediately after sunset" with a continuous and smooth function, avoiding the interference of "sudden changes" of light weight on system decision-making;

[0079] In traditional designs, environmental parameter weights are either determined entirely by the environment itself (e.g., higher weights for higher temperatures) or fixed by user commands (e.g., "turn off all controls when away from home"). However, these formulas utilize user mode (m) as the "centerpiece of environmental weighting," enabling dynamic adaptation between environmental factors and human needs. Temperature's impact on the system depends not only on its physical value but also on the user's state (e.g., a higher tolerance for temperature fluctuations during sleep). This solution breaks with the traditional logic of "temperature weighting determined solely by temperature value" and for the first time incorporates "human subjective state" as a "modulating variable" for environmental parameter weights, enabling dynamic temperature weighting based on user needs. This avoids the logical confusion of the traditional approach of lumping all environmental parameters together and adjusting their weights.

[0080] In one embodiment of the present invention, the first adjustment coefficient vector includes:

[0081] The first sub-coefficient is used to adjust the basic weight of outdoor temperature and humidity;

[0082] The second sub-coefficient is used to adjust the basic weight of light intensity;

[0083] The second adjustment coefficient vector includes:

[0084] The third sub-coefficient is used to adjust the basic weight of outdoor temperature and humidity;

[0085] The fourth sub-coefficient is used to adjust the basic weight of PM2.5 concentration;

[0086] The fifth sub-coefficient is used to adjust the basic weight of weather forecast data.

[0087] The first sub-coefficient is (f(t)×g(t)), and the second sub-coefficient is ), the third sub-coefficient is , the fourth sub-coefficient is , the fifth sub-coefficient is .

[0088] The working principle and effect of the above technical solution are as follows: In an embodiment of the present invention, the first adjustment coefficient vector dynamically adjusts the basic weights of outdoor temperature and humidity and light intensity through the first sub-coefficient and the second sub-coefficient, wherein the first sub-coefficient is generated based on the time parameter, and the second sub-coefficient is generated based on the sunrise and sunset time through the sigmoid function; the second adjustment coefficient vector adjusts the basic weights of outdoor temperature and humidity, PM2.5 concentration, and weather forecast data through the third sub-coefficient, the fourth sub-coefficient, and the fifth sub-coefficient, wherein the third sub-coefficient is generated based on the user's preset mode, the fourth sub-coefficient is associated with the user's health concern level, and the fifth sub-coefficient is combined with the user's safety needs; the first and second adjustment coefficient vectors have clear division of labor, the former adjusts the weights for temperature, humidity, and light in combination with natural laws, and the latter adjusts the weights for temperature, humidity, PM2.5, and meteorological data in combination with user modes and health and safety needs, working synergistically and normalizing the weights to avoid the dominance of a single parameter and improve decision-making accuracy and situational adaptability.

[0089] In one embodiment of the present invention, the S4 specifically includes:

[0090] The fuzzy variables of each environmental parameter and the corresponding environmental parameter weight vector are fused to generate an intermediate fuzzy state; and the weight of each environmental parameter is Based on the inference rules in the intermediate fuzzy state matching rule library, the corresponding conclusion fuzzy subset is triggered, and the triggered conclusion fuzzy subset is aggregated. The outputs of multiple rules are integrated through weighted summation or the larger or smaller method to generate comprehensive fuzzy evaluation values ​​of window opening suitability, window closing urgency, and shading demand.

[0091] The working principle and effect of the above technical solution are as follows: first, the fuzzy variables of each environmental parameter (the fuzziness of "hot" for temperature is 0.7, the fuzziness of "light pollution" for PM2.5 is 0.6, the fuzziness of "medium probability rain" for rainfall probability is 0.5, etc.) are fused with their corresponding environmental parameter weight vectors (temperature weight 0.3, PM2.5 weight 0.2, rainfall probability weight 0.2, etc.), and the fuzzy fuzziness of each parameter is multiplied by the corresponding weight and superimposed to generate an intermediate fuzzy state. Subsequently, based on this intermediate fuzzy state, the inference rules in the rule library are matched ("If the temperature is hot (fuzziness ≥ 0.5) and the PM2.5 is good (fuzziness ≥ 0.6), then the window opening suitability is high (fuzziness 0.8)" "If the rainfall probability is medium probability rain (fuzziness ≥ 0.4) and the wind speed is moderate (fuzziness ≥ 0.5), then the window closing urgency is medium (fuzziness 0.6)" etc.), triggering the corresponding conclusion fuzzy subset; then, perform aggregation operation on the triggered multiple conclusion fuzzy subsets. If the weighted summation method is adopted, weights are assigned according to the matching degree of each rule (the degree of fit between the intermediate state and the rule premise) and then accumulated (Rule A matching degree 0.7 outputs window opening suitability 0.8, Rule B matching degree 0.5 outputs window opening suitability 0.3, comprehensive value = 0.7×0.8+0.5×0.3=0.71). If the larger or smaller method is adopted, the maximum or minimum value output by each rule is taken (the urgency of closing the window takes the maximum value of the rule output 0.6), and finally a comprehensive fuzzy evaluation value of window opening suitability, window closing urgency, and shading demand is generated; when there is a rule matching conflict (Rule 1 recommends "window opening suitability 0.7" and Rule 2 recommends "window opening suitability 0.7" and Rule 3 recommends "window closing suitability 0.3", comprehensive value = 0.7×0.8+0.5×0.3=0.71). When the system recommends "window closing urgency 0.8"), the system calls the conflict resolution rule group and adjusts the weight of the conflicting rules based on the current situation parameters (quietness is prioritized in Sleep mode, and safety is prioritized in Away mode) (the weight of the window closing rule is increased to 0.7 in Sleep mode, and the weight of the window opening rule is reduced to 0.3);

[0092] By integrating the fuzzy variables of environmental parameters and the corresponding weights, the one-sidedness of being dominated by a single parameter is avoided; complex scenarios of multi-parameter collaboration are accurately matched based on the intermediate fuzzy state to fit the actual working conditions; when aggregating multiple rule outputs, the decision-making basis is integrated through weighting or extreme value method to avoid extremism; when rules conflict, the weights of conflicting rules are dynamically adjusted according to the situational parameters, giving priority to meeting core needs and resolving contradictions; and the overall decision-making accuracy, stability and situational adaptability are improved.

[0093] One embodiment of the present invention provides an intelligent door and window control system, the system comprising:

[0094] Multi-source environmental perception system: obtains multi-source environmental data;

[0095] Fuzzy variable conversion system: multi-source environmental data is input into the fuzzification module and converted into fuzzy variables through the preset fuzziness function;

[0096] Dynamic weight adaptation system: dynamically generates environmental parameter weight vectors based on real-time scenario parameters;

[0097] Fuzzy reasoning decision system: Fuzzy variables and environmental parameter weight vectors are input into the fuzzy reasoning engine, logical operations of the fuzzy rule base are executed, and comprehensive evaluation values ​​of window opening suitability, window closing urgency, and shading demand are output;

[0098] Control instruction execution system: Generates the final control instruction through defuzzification strategy and executes it.

[0099] The working principle and effect of the above technical solution are as follows: the system first obtains multi-source environmental data such as outdoor temperature and humidity, light intensity, PM2.5 concentration, and weather forecast through sensors and meteorological APIs. This data is then input into the fuzzification module and converted into fuzzy variables that conform to natural language descriptions using a preset trapezoidal fuzziness function. This replaces the hard threshold to avoid decision-making oscillations caused by boundary fluctuations. At the same time, the system dynamically generates environmental parameter weight vectors based on real-time scenario parameters. The time parameter adjusts the weights of temperature, humidity, and light through a periodic function and a seasonal amplifier. The user mode adjusts the weights of temperature, humidity, PM2.5, and meteorological data through a mode-specific correction factor. The meteorological data confidence factor further corrects the weights. Afterwards, the fuzzy inference engine inputs the above fuzzy variables and dynamic weight vectors into a preset rule library. By matching multiple sets of condition-conclusion inference rules, logical operations are performed to output comprehensive fuzzy evaluation values ​​for window opening suitability, window closing urgency, and shading demand. Finally, the comprehensive evaluation values ​​are converted into specific control instructions through a defuzzification strategy and executed, realizing the processing of data conflicts and fuzzy boundaries.

[0100] It solves the problem of frequent invalid operations caused by fluctuations in environmental parameters near boundary values, and improves the smoothness and stability of decision-making; by dynamically generating environmental parameter weight vectors based on time parameters, user preset modes and meteorological data confidence factors, it achieves adaptation to scenarios such as seasons, time periods, and user status, so that the decision logic can be flexibly adjusted according to actual needs; it avoids contradictory decisions under traditional fixed rules, and improves the rationality of decisions in complex environments.

[0101] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A smart door and window control method, characterized in that: The method comprises: S1: Acquire multi-source environmental data; S2: Input the multi-source environmental data into the fuzzification module and convert it into fuzzy variables through the preset fuzziness function; S3: Dynamically generate environmental parameter weight vectors based on real-time scenario parameters; S4: Input the fuzzy variables and environmental parameter weight vectors into the fuzzy inference engine, execute the logical operation of the fuzzy rule base, and output the comprehensive evaluation values ​​of window opening suitability, window closing urgency, and shading demand; S5: Generate the final control instructions through the defuzzification strategy and execute them.

2. The intelligent door and window control method according to claim 1, characterized in that: The multi-source environmental data of S1 includes outdoor temperature and humidity, light intensity, PM2.5 concentration and weather forecast data.

3. The intelligent door and window control method according to claim 1, characterized in that: The fuzziness function adopts a trapezoidal function, and defines multi-source environmental data including multiple gradient levels, wherein: The fuzziness functions of temperature, humidity, and light dynamically adjust boundary parameters according to the season; The fuzziness function of PM2.5 concentration is associated with the health standard threshold.

4. The intelligent door and window control method according to claim 2, characterized in that: The weather forecast data is processed in the following manner: Generate confidence factors based on the update timeliness of meteorological data and the granularity of meteorological data forecast; Convert the predicted rainfall probability and wind speed level into fuzzy variables: Rainfall probability gradient: the first gradient is no rain, the second gradient is medium probability rain, and the third gradient is high probability rain; Wind speed level gradient: first level is light wind, second level is moderate wind, and third level is strong wind.

5. The intelligent door and window control method according to claim 1, characterized in that: The real-time scenario parameters of S3 include: Acquire real-time scenario parameters, including: Time parameters and user preset modes, wherein the user preset modes include: sleep mode and away mode.

6. The intelligent door and window control method according to claim 1, characterized in that: The environmental parameter weight vector is obtained as follows: The preset multi-source environmental data includes weights of outdoor temperature and humidity, light intensity, PM2.5 concentration, and weather forecast data, and the sum of the weights of the multi-source environmental data is 1; generating a first adjustment coefficient vector based on the time parameter, for adjusting the weights of outdoor temperature and humidity and light intensity; Generate a second adjustment coefficient vector based on a user preset mode, for adjusting the weights of outdoor temperature and humidity, PM2.5 concentration, and weather forecast data; A third adjustment coefficient vector is generated based on the confidence factor, and is used to adjust the weight of the weather forecast data.

7. The intelligent door and window control method according to claim 6, characterized in that: The first adjustment coefficient vector includes: The first sub-coefficient is used to adjust the basic weight of outdoor temperature and humidity; The second sub-coefficient is used to adjust the basic weight of light intensity; The second adjustment coefficient vector includes: The third sub-coefficient is used to adjust the basic weight of outdoor temperature and humidity; The fourth sub-coefficient is used to adjust the basic weight of PM2.5 concentration; The fifth sub-coefficient is used to adjust the basic weight of weather forecast data.

8. The intelligent door and window control method according to claim 1, characterized in that: The S4 specifically includes: The fuzzy variables of each environmental parameter and the corresponding environmental parameter weight vector are fused to generate an intermediate fuzzy state; Based on the inference rules in the intermediate fuzzy state matching rule library, the corresponding conclusion fuzzy subset is triggered, and the triggered conclusion fuzzy subset is aggregated. The outputs of multiple rules are integrated through weighted summation or the larger or smaller method to generate comprehensive fuzzy evaluation values ​​of window opening suitability, window closing urgency, and shading demand.

9. The intelligent door and window control method according to claim 8, characterized in that: The rule base also includes: When a rule matching conflict occurs, the conflict resolution rule group is called, the weight ratio of the conflicting rules is adjusted according to the current situation parameters, and instructions that meet the situation priority are generated first.

10. An intelligent door and window control system, characterized in that: The system comprises: Multi-source environmental perception system: obtains multi-source environmental data; Fuzzy variable conversion system: multi-source environmental data is input into the fuzzification module and converted into fuzzy variables through the preset fuzziness function; Dynamic weight adaptation system: dynamically generates environmental parameter weight vectors based on real-time scenario parameters; Fuzzy reasoning decision system: Fuzzy variables and environmental parameter weight vectors are input into the fuzzy reasoning engine, logical operations of the fuzzy rule base are executed, and comprehensive evaluation values ​​of window opening suitability, window closing urgency, and shading demand are output; Control instruction execution system: Generates the final control instruction through defuzzification strategy and executes it.

Images