Air conditioner refrigeration energy-saving self-adaptive control method based on artificial intelligence

Abstract

The invention relates to the technical field of air conditioner control, and discloses an air conditioner refrigeration energy-saving self-adaptive control method based on artificial intelligence, which comprises the following steps: collecting indoor member activity tracks and indoor temperature data, analyzing habitual environment temperatures and habitual sensible temperatures of members, simulating indoor temperature distribution under different operating parameters of an air conditioner, and calculating the indoor temperature distribution under different operating parameters of the air conditioner; according to the air conditioner refrigeration energy-saving self-adaptive control method based on artificial intelligence, personalized temperature adjustment is provided by analyzing the habitual environment temperature and the sensible temperature of the members, the comfortable experience of the members is improved, and the air conditioner refrigeration energy-saving self-adaptive control method based on artificial intelligence has the advantages that the air conditioner refrigeration energy-saving self-adaptive control method based on artificial intelligence is improved by optimizing the air conditioner parameters; unnecessary energy consumption is reduced, the energy utilization efficiency is improved, automatic and intelligent management of the air conditioning system is achieved, manual intervention is reduced, the management efficiency is improved, and the system can dynamically adjust air conditioning parameters according to indoor and outdoor environment changes and movement tracks of members, so that it is ensured that the indoor temperature is always in the optimal state.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of air conditioner control, in particular to an air conditioner refrigeration energy-saving self-adaptive control method based on artificial intelligence. BACKGROUND

[0002] The basic working principle of a traditional air conditioner controller (such as a simple temperature controller) is to maintain a fixed target temperature, and when the indoor temperature deviates from the set value by a certain range, the compressor is started or stopped. This "on-off" control is very rough, although the variable frequency air conditioner is more energy-saving than the fixed frequency air conditioner, but its fine regulation and control capability is often limited by the simple thermostat logic, and the optimal flexible operation cannot be realized. The core of the air conditioner refrigeration energy-saving self-adaptive control method based on artificial intelligence is to use artificial intelligence technology to learn and analyze the use habits of users, indoor environment data (temperature, humidity, illumination, number of occupants, etc.), outdoor environment data (temperature, humidity, illumination, etc.) and the running state of the air conditioner itself, to predict the needs of users and optimize the operation strategy of the air conditioner, so as to achieve the balance between energy saving and comfort. It is a deep learning and prediction of environment, users and equipment by using AI technology, so as to realize the optimal energy utilization and dynamic response of comfort, which represents the development direction of future smart home and energy-saving technology.

[0003] The existing air conditioner refrigeration energy-saving self-adaptive control method based on artificial intelligence cannot analyze the use habits of each user in the room, especially the habit temperature, cannot analyze the temperature difference of different positions in the room, especially the temperature influence on each position in the room after the air conditioner is running, cannot comprehensively analyze the habits of all users in the same space, combine the outdoor temperature, indoor natural wind force, indoor layout and the influence of the air output after the air conditioner is running on each position in the room, and adjust to a suitable temperature to make all users in the same space feel comfortable as much as possible. It is easy to cause the influence of different environmental temperatures and different body temperatures of users in the same space, so that the user experience is inconsistent, such as one user feels cool while another user feels cold. The practicality has certain limitations. SUMMARY

[0004] The application provides an air conditioner refrigeration energy-saving self-adaptive control method based on artificial intelligence, which is used for promoting the solution to the problems in the background technology.

[0005] The application provides the following technical scheme: an air conditioner refrigeration energy-saving self-adaptive control method based on artificial intelligence, comprising:

[0006] Collecting the activity track of the indoor member;

[0007] For each indoor member, the activity track is extracted and defined as a member track.

[0008] Collecting indoor temperature data;

[0009] For each member trajectory, extract its data point set, and define it as a member trajectory point set;

[0010] For each data point in the member trajectory point set, simulate the environmental temperature of each data point under different refrigeration durations to obtain the refrigeration environmental temperature;

[0011] For each data point in the member trajectory point set, analyze the stay environmental temperature of each data point;

[0012] For each indoor member, analyze the habit environmental temperature and habit sensible temperature of the member;

[0013] In combination with the current indoor and outdoor environment and the member habits, simulate the indoor temperature distribution under different operation parameters of the air conditioner;

[0014] Determine the final air conditioner refrigeration temperature, air supply angle and air speed;

[0015] According to the final air conditioner refrigeration temperature, air supply angle and air speed, control the air conditioner to operate.

[0016] As an optional solution of the air conditioner refrigeration energy-saving adaptive control method based on artificial intelligence, the indoor member activity trajectory is collected, and the collection specifically includes:

[0017] All indoor members are obtained;

[0018] An indoor coordinate system is defined;

[0019] A fixed time interval is set;

[0020] For each indoor member, record the position coordinates of the member at different time points by using a time sequence to form a trajectory sequence:

[0021] As an optional solution of the air conditioner refrigeration energy-saving adaptive control method based on artificial intelligence, the indoor temperature data is collected, and the collection specifically includes:

[0022] Air conditioner refrigeration data of each indoor member when controlling the air conditioner is obtained, and a parameter sequence is formed:

[0023]

[0024] All temperature acquisition devices in the indoor are obtained: J = (j1, j2,..., j n );

[0025] For each temperature acquisition device, obtain the device position and effective acquisition range thereof, and denote them as and

[0026] extracting a member trajectory of each indoor member;

[0027] setting a plurality of data points on the member trajectory to form a data point set;

[0028] For each data point p l (x l , y l ) in the data point set, calculate the distance between it and each temperature collection device:

[0029] determine the regional ambient temperature of each data point in the data point set:

[0030] define a temperature value analysis function to determine the actual ambient temperature of each data point in the data point set:

[0031] if F R_T = 1, the regional ambient temperature of the data point is identified as the actual ambient temperature of the data point;

[0032] if F R_T = 0, the actual ambient temperature of the data point is determined by speculation.

[0033] As an optional solution of the air conditioning refrigeration energy-saving adaptive control method based on artificial intelligence, the actual ambient temperature of the data point is determined by speculation, specifically:

[0034] obtain the natural wind direction and natural wind speed of the data point;

[0035] calculate the temperature influence of natural wind, and define it as natural wind temperature:

[0036] obtain the regional ambient temperature of the data point;

[0037] calculate the weight coefficient of the regional ambient temperature and the natural wind temperature:

[0038]

[0039] combine the regional ambient temperature and the natural wind temperature to obtain the actual ambient temperature:

[0040] T estimated = α·T region +(1-α)·T natural .

[0041] As an optional solution of the air conditioning refrigeration energy-saving adaptive control method based on artificial intelligence, the actual ambient temperature of the data point is determined by simulation under different refrigeration time, and the refrigeration ambient temperature is obtained, specifically:

[0042] Obtain the actual environment temperature of the i th data point at t-1 time, as the current environment temperature, denoted as T real，i (t-1);

[0043] Calculate the reduction value of the air conditioner on the environment temperature at the current time: ΔT air =(T aircond -T real，i (t-1))·α(t);

[0044] Wherein, α(t) is the attenuation coefficient of refrigeration effect, the specific formula is:

[0045]

[0046] Calculate the influence value of natural wind on the environment temperature at the current time:

[0047] ΔT wind =v wind ·cos(θ)·(T outdoor -T real，i (t-1))·k wind ;

[0048] Add the initial environment temperature, air conditioning refrigeration effect and natural wind influence to obtain the actual environment temperature of the i th data point at t time, as the refrigeration environment temperature, denoted as T real，i (t):

[0049] T real，i (t)=T real，i (t-1)+ΔT air +ΔT wind .

[0050] As an optional solution of the air conditioning refrigeration energy-saving self-adaptive control method based on artificial intelligence, wherein: analyze the stay environment temperature of each data point, specifically:

[0051] Calculate the position change of adjacent two time points: D i，t =|x i，t+1 -x i，t | 2 +|y i，t+1 -y i，t | 2 ;

[0052] Define a state judgment function to judge whether the member is in a stay state within adjacent two time points:

[0053] If Status i，t =1, it is judged that the member is in a stay state within adjacent two time points;

[0054] If Status i，t = 0, it is determined that the member is in a moving state within two adjacent time points;

[0055] A continuous state analysis function is defined to determine whether the member is in a stay state at the data point:

[0056]

[0057] If Status_continuous i，t = 1, it is determined that the member is in a stay state at the data point;

[0058] If Status_continuous i，t = 0, it is determined that the member is in a moving state at the data point;

[0059] In the determined stay period, the refrigeration environment temperature T real，i (t) corresponding to each time point is obtained;

[0060] The average value of the temperature values in the stay period is calculated:

[0061] The weighted stay environment temperature is calculated:

[0062] As an optional solution of the air conditioner refrigeration energy-saving self-adaptive control method based on artificial intelligence, wherein: the habit environment temperature and the habit body temperature of the member are analyzed, specifically:

[0063] For each indoor member, the member trajectory thereof is extracted;

[0064] For each data point on the member trajectory, the stay duration of the member at the data point is recorded, denoted as t i ;

[0065] The stay environment temperature T stay，i at the data point is obtained;

[0066] The habit environment temperature of the member is calculated:

[0067] The outdoor environment temperature is obtained, denoted as T outdoor ;

[0068] The natural wind speed in the room is obtained, denoted as v wind ;

[0069] For each data point, the body temperature influence factor is calculated:

[0070] Impact_Factori = T stay，i + k1·T outdoor + k2·v wind ;

[0071] Calculate the habit temperature of the member:

[0072] As an optional solution of the air conditioning refrigeration energy-saving adaptive control method based on artificial intelligence, wherein: in combination with the current indoor and outdoor environment and the member habits, the indoor temperature distribution under different operating parameters of the air conditioner is simulated, specifically:

[0073] Obtain the initial temperature of the indoor;

[0074] Obtain the outdoor temperature T outdoor ;

[0075] Obtain the natural wind speed and direction of the indoor;

[0076] Obtain all the refrigeration temperatures, air supply angles and air supply speeds of the air conditioner;

[0077] Map each air supply angle of the air conditioner to each air supply speed one by one to form an air supply force set;

[0078] Map each element in the air supply force set to each refrigeration temperature one by one to form a simulated air supply set, denoted as P;

[0079] For each set of air conditioner parameter combinations in the simulated air supply set, simulate the air supply of the air conditioner to the indoor;

[0080] Set the time point of the simulated air supply of the air conditioner to the indoor as the simulation start time;

[0081] Set an interval analysis duration;

[0082] Take the simulation start time as the starting point, and form a time collection point every interval analysis duration;

[0083] Calculate the indoor temperature at each time collection point:

[0084]

[0085] After the air conditioner operates for different lengths of time under each set of air conditioner parameter combinations, the indoor temperature T real，i (t) of each location in the indoor at each time point is stored in a data structure to obtain the simulation data of the indoor temperature distribution.

[0086] As an optional solution of the air conditioning refrigeration energy-saving adaptive control method based on artificial intelligence, wherein: determine the final air conditioning refrigeration temperature, air supply angle and air speed, specifically:

[0087] For each data point on each member's trajectory, record the duration t of that member's stay at that data point. i ;

[0088] Calculate the weighted value of temperature change and dwell time for each data point:

[0089]

[0090] Calculate the weight of the positional overlap between each data point and the member's activity trajectory:

[0091] The weighted temperature change of each data point is combined with the location overlap weight to obtain the comprehensive temperature change.

[0092] Calculate the weighted average temperature T of the simulation based on the duration of each member's stay at each data point. simulated :

[0093]

[0094] Calculate the simulated temperature weighted average T, taking into account the degree of positional overlap. simulated_overlap :

[0095]

[0096] Calculate the weighted average value T of the simulated temperature. simulated With the members' habitual ambient temperature T habit Error simulated Error simulated =|T simulated -T habit |;

[0097] Calculate the simulated temperature weighted average T, taking into account the degree of positional overlap. simulated_overlap With the members' habitual ambient temperature T habit Error simulated_overlap Error simulated_overlap =|T simulated_overlap -T habit |;

[0098] Calculate the comprehensive error assessment value Error_total: Error_total = Error simulated Error simulated_overlap ;

[0099] Iterate through all combinations of air conditioning parameters in the simulated air supply set P, and record the combination with the smallest error. This combination is the optimal air conditioning parameter combination that best matches the ambient temperature preferred by the members.

[0100] The air conditioning refrigeration temperature, the air supply angle and the air speed in the optimal air conditioning parameter combination are the final air conditioning refrigeration temperature, the air supply angle and the air speed.

[0101] The present application has the following advantages:

[0102] 1. The air conditioning refrigeration energy-saving self-adaptive control method based on artificial intelligence collects the historical activity trajectory of the indoor member, records the position coordinates and time stamp of the member at different time points, collects the air conditioning setting parameters, records the refrigeration temperature, air supply angle and air supply speed set by the member when controlling the air conditioner, collects the environmental temperature of each position in the room, uses the temperature collection equipment to obtain the actual environmental temperature of each position in the room, determines the position and effective collection range of the temperature collection equipment, calculates the distance between each data point and the equipment, so as to obtain the most accurate regional environmental temperature, and through the collection of the historical activity trajectory of the member, the air conditioning setting parameters and the indoor environmental temperature, a comprehensive data basis is provided for subsequent analysis and optimization, the position and effective collection range of the temperature collection equipment are determined, and the accuracy and reliability of the obtained environmental temperature data are ensured.

[0103] 2. The air conditioning refrigeration energy-saving self-adaptive control method based on artificial intelligence analyzes the habit environmental temperature of the member, calculates the habit environmental temperature of the member through the staying time of the member at different data points and the air conditioning setting, combined with the environmental temperature change, analyzes the thermal sensation of the member, combined with the outdoor environmental temperature, natural wind direction and speed and indoor layout, simulates the indoor temperature change, determines the actual environmental temperature of each position, calculates the habit thermal sensation of the member, according to the staying time of the member at different data points and the thermal sensation, calculates the habit thermal sensation of the member, through the analysis of the habit environmental temperature and the thermal sensation of the member, the individualized temperature demand of the member can be accurately reflected, combined with the outdoor environmental temperature, natural wind direction and speed and indoor layout, the indoor temperature change is simulated, and dynamic temperature distribution data is provided.

[0104] 3. This AI-based adaptive control method for air conditioning cooling energy saving simulates indoor temperature distribution under different air conditioning operating parameters. Using a dynamic temperature change formula, it simulates indoor temperature distribution under different combinations of air conditioning parameters, calculates the weighted average of the simulated temperature, and calculates the weighted average of the simulated temperature based on the duration of each member's stay at each data point. It introduces a positional overlap weight, calculating the overlap weight between each data point and the member's activity trajectory, comprehensively considering the positional overlap of the simulated temperature weighted average, comparing the error, calculating the error between the simulated temperature and the member's preferred ambient temperature, selecting the air conditioning parameter combination with the smallest error, and controlling the air conditioning to operate with the selected parameters based on the optimization results to ensure the indoor temperature meets the member's comfort needs. By simulating indoor temperature distribution under different air conditioning parameter combinations, it selects the parameter combination that best matches the member's preferred ambient temperature, achieving precise temperature regulation. Introducing a positional overlap weight, it comprehensively considers the member's stay time and temperature changes at different locations, providing a more comprehensive error assessment, and automatically adjusting the air conditioning parameters based on the optimization results to ensure the indoor temperature always meets the member's comfort needs. By dynamically adjusting the air conditioning parameters, it improves the comfort of the indoor environment and enhances the member's quality of life. Attached Figure Description

[0105] Figure 1 This is a flowchart of the air conditioning cooling energy-saving adaptive control method based on artificial intelligence according to the present invention. Detailed Implementation

[0106] 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.

[0107] Example 1: An AI-based adaptive energy-saving control method for air conditioning refrigeration (see...) Figure 1 ,include:

[0108] Collect the activity trajectories of indoor occupants;

[0109] For each indoor member, their activity trajectory is extracted and defined as the member trajectory;

[0110] Collect indoor temperature data;

[0111] For each member trajectory, extract its set of data points and define it as the member trajectory point set;

[0112] For each data point in the set of member trajectory points, simulate the ambient temperature of each data point under different cooling durations to obtain the cooling ambient temperature;

[0113] For each data point in the member trajectory point set, analyze the stay environment temperature of each data point;

[0114] For each indoor member, analyze the habit environment temperature and the habit thermal sensation temperature of the member;

[0115] In combination with the current indoor and outdoor environment and the member habit, simulate the indoor temperature distribution under different operation parameters of the air conditioner;

[0116] Determine the final air conditioner refrigeration temperature, air supply angle and air speed;

[0117] According to the final air conditioner refrigeration temperature, air supply angle and air speed, control the air conditioner to operate.

[0118] Through the above method, by analyzing the habit environment temperature and the thermal sensation temperature of the member, personalized temperature adjustment is provided, the comfort experience of the member is improved, unnecessary energy consumption is reduced by optimizing the air conditioner parameters, energy utilization efficiency is improved, automatic and intelligent management of the air conditioner system is realized, manual intervention is reduced, management efficiency is improved, the system can dynamically adjust the air conditioner parameters according to the indoor and outdoor environment changes and the activity trajectory of the member, and ensure that the indoor temperature is always in the best state.

[0119] Embodiment two, this embodiment is an improvement based on embodiment one, the air conditioner refrigeration energy-saving self-adaptive control method based on artificial intelligence, collects the activity trajectory of the indoor member, specifically:

[0120] Get all indoor members, the indoor members are permanent residents in the space;

[0121] Define an indoor coordinate system to ensure that each position can be represented by a unique (x, y) coordinate;

[0122] Set a fixed time interval, such as a fixed time interval of 5 seconds;

[0123] For each indoor member, record its position coordinates at different time points in a time sequence to form a trajectory sequence:

[0124]

[0125] Wherein, i is the member number, t is the recording number, the recording number +1 every fixed time interval, T is the time sequence length, the time sequence length T refers to the total number of times of recording member position information under a fixed time interval (such as 5 seconds), specifically For example, if a person is recorded to be active in a room for 1 minute (60 seconds), and the position is recorded every 5 seconds, then T = (60 / 5) + 1 = 13 records, which can accurately describe the length of the time series, and the total length of time in the time series T is 3 months.

[0126] The embodiment also provides that the indoor temperature data is collected, in particular:

[0127] The air conditioner refrigeration data when each indoor member controls the air conditioner is acquired, and a parameter sequence is formed:

[0128]

[0129] wherein, T aircond，k is the refrigeration temperature when the indoor member controls the air conditioner, θ k is the air supply angle when the indoor member controls the air conditioner, v k is the air supply speed when the indoor member controls the air conditioner, K is the number of indoor members, and k is the number of each indoor member.

[0130] All temperature acquisition devices in the room are acquired, the temperature acquisition devices are all devices that can acquire the ambient temperature in the indoor space, including air conditioners, smart home devices, etc.

[0131] J = (j1, j2,..., j n );

[0132] For each temperature acquisition device, the device position and effective acquisition range thereof are acquired, and are denoted as and

[0133] For each indoor member, the member trajectory thereof is extracted.

[0134] A plurality of data points are set on the member trajectory to form a data point set.

[0135] For each data point p l (x l , y l ) in the data point set, the distance between the data point and each temperature acquisition device is calculated.

[0136]

[0137] The regional ambient temperature of each data point in the data point set is determined.

[0138]

[0139] wherein, T_min(·) is a temperature position function, i.e. the temperature value of the temperature collection device closest to the data point;

[0140] defining a temperature value analysis function to determine the actual environmental temperature of each data point in the data point set:

[0141]

[0142] wherein, is the distance threshold of the temperature collection device collecting the regional environmental temperature of the data point, i.e. the radius value of the effective collection range of the temperature collection device closest to the data point, and min(·) is a position function, i.e. the distance value of the temperature collection device closest to the data point;

[0143] if F R_T = 1, the regional environmental temperature of the data point is identified as the actual environmental temperature of the data point;

[0144] if F R_T = 0, the actual environmental temperature of the data point is determined by inference;

[0145] wherein, a plurality of data points are set on the member trajectory to form a data point set, specifically:

[0146] collecting the stay points of the indoor member on the member trajectory:

[0147] P_ST = Collect_ST{sp(x, y) | sp(t) ≥ t threshold};

[0148] wherein, t threshold is a stay threshold for determining whether the member stays at the location for a long time, sp(t) is the duration of the member staying at the location, sp(x, y) is the position coordinates of the location where the member stays for a long time, and Collect_ST{·} is a stay point function for collecting the position coordinates of the location where the member stays for a long time, i.e. collecting the stay points of the member;

[0149] successively extracting two adjacent stay points, and defining the trajectory between the two adjacent stay points as an adjacent trajectory segment;

[0150] uniformly setting a plurality of division points in the adjacent trajectory segment:

[0151]

[0152] wherein, trajectory is the adjacent trajectory segment, D preset is the preset distance between two adjacent division points, indicates that the adjacent track segment can divide the distance between the two adjacent preset segmentation points evenly, indicates that the adjacent track segment cannot divide the distance between the two adjacent preset segmentation points evenly, n preset is the number of preset segmentation points, which is an integer obtained by rounding off the value obtained by dividing the adjacent track segment by the distance between the two adjacent preset segmentation points, Segmentation(·, ·) is a segmentation function for evenly segmenting the adjacent track segment into a plurality of segmentation points according to the distance between the two adjacent preset segmentation points or the number of preset segmentation points, is a segmentation point collection function for collecting all segmentation points of the member, that is, when the adjacent track segment can divide the distance between the two adjacent preset segmentation points evenly, a plurality of segmentation points are evenly arranged in the adjacent track segment according to the distance between the two adjacent preset segmentation points, and when the adjacent track segment cannot divide the distance between the two adjacent preset segmentation points evenly, a plurality of segmentation points are evenly arranged in the adjacent track segment according to the number of preset segmentation points;

[0153] All stay points and segmentation points on all member tracks are integrated to generate a data point set:

[0154] P_D={P_ST∪P_SE}={p1(x1, y1), p2(x2, y2),..., p n (x n , y n )}.

[0155] wherein the actual ambient temperature of the data point is determined, specifically:

[0156] The natural wind direction and natural wind speed of the data point are obtained;

[0157] The temperature influence of the natural wind is calculated and defined as the natural wind temperature:

[0158]

[0159] wherein T outdoor is the outdoor temperature, v wind is the natural wind speed of the data point, w1 is the influence weight of the natural wind direction on the temperature, which is used to measure the contribution degree of the natural wind to the temperature change under different wind directions, and the specific formula is a is an adjustment parameter, which is used to describe the sensitivity of the wind direction and wind speed to the weight, θ is the wind direction of the natural wind, and w2 is the influence weight of the natural wind speed on the temperature, which is used to measure the contribution degree of the natural wind to the temperature change under different wind speeds, and the specific formula is b is an adjustment parameter, which is used to describe the sensitivity of the wind direction and wind speed to the weight, and these weights will affect the temperature change brought by the natural wind, and further affect the calculation of the indoor environment temperature;

[0160] Obtain the regional ambient temperature of the data point;

[0161] Calculate the weight coefficient of the regional ambient temperature and the natural wind temperature:

[0162]

[0163] Wherein, k is the wind speed influence coefficient, used to adjust the influence degree of wind speed on the weight of regional ambient temperature and natural wind temperature, the sign of k determines how wind speed affects the weight, positive number makes the weight of regional ambient temperature increase with the increase of wind speed, and negative number is the opposite, the absolute value of k reflects the sensitivity of wind speed to the change of weight, the larger the absolute value, the more significant the influence of wind speed change on the weight;

[0164] Combine the regional ambient temperature and the natural wind temperature to obtain the actual ambient temperature:

[0165] T estimated = α·T region + (1-α)·T natural ;

[0166] Wherein, T region is the regional ambient temperature of the data point.

[0167] Example three, this embodiment is improved on the basis of example two, in this embodiment, the ambient temperature of each data point under different refrigeration time is simulated to obtain the refrigeration ambient temperature, specifically:

[0168] Obtain the actual ambient temperature of the i-th data point at t-1 moment, as the current ambient temperature, denoted as T real，i (t-1);

[0169] Calculate the reduction value of the air conditioner to the ambient temperature at the current moment:

[0170] ΔT air = (T aircond - T real，i (t-1))·α(t);

[0171] Wherein, α(t) is the attenuation coefficient of refrigeration effect, the specific formula is:

[0172]

[0173] Wherein, T aircond is the air conditioner refrigeration temperature, T outdooris the outdoor temperature, kk and c are coefficients for adjusting the cooling effect, kk mainly affects the degree of attenuation of the cooling effect over time, the larger the value of kk, the faster the cooling effect attenuates over time, c adjusts the influence of the indoor and outdoor temperature difference on the cooling effect, the larger the value of c, the more sensitive the cooling effect is to the temperature difference, that is, when the temperature difference is slightly large, the cooling effect will be significantly enhanced, and when the temperature difference is small, the cooling effect will rapidly weaken, and the two coefficients together determine the trend of the cooling effect over time and the temperature difference;

[0174] Calculate the influence value of natural wind on the ambient temperature at the current time:

[0175] ΔT wind = v wind · cos (θ) · (T outdoor - T real，i (t-1)) · k wind ;

[0176] Where k wind is the adjustment coefficient of natural wind speed, reflecting the influence of wind speed on heat exchange efficiency, which is usually adjusted according to the actual environment and equipment characteristics, v wind is the natural wind speed, θ is the angle between the wind direction and the data point, and cos (θ) is used to directionally weight the wind speed, when the wind direction is consistent with the direction of the environment point, the influence is the largest;

[0177] Add the initial ambient temperature, air conditioning cooling effect and natural wind influence to obtain the actual ambient temperature of the ith data point at time t, which is defined as the cooling ambient temperature, denoted as T real，i (t):

[0178] T real，i (t) = T real，i (t-1) + ΔT air + ΔT wind .

[0179] The embodiment also provides that the stay ambient temperature of each data point is analyzed, specifically:

[0180] Calculate the position change of adjacent two time points:

[0181] D i，t = |x i，t+1 - x i，t | 2 + |y i，t+1 - y i，t | 2 ;

[0182] Where x i，t and y i，t represent the position coordinates of member i at time point t, which are used to record the activity trajectory of the member at different positions in the room, D i，trepresents the square of the distance moved by member i between time points t and t+1, used to determine whether the member is moving;

[0183] A state determination function is defined to determine whether the member is in a stationary state between adjacent time points:

[0184]

[0185] wherein δ stay is the threshold value of the stationary determination, used to determine whether the member is in a stationary state within the time period, if D i，t is less than δ stay , it is considered that the member is in a stationary state within the time period, otherwise, it is considered that the member is in a moving state within the time period;

[0186] If Status i，t = 1, it is determined that the member is in a stationary state between adjacent time points;

[0187] If Status i，t = 0, it is determined that the member is in a moving state between adjacent time points;

[0188] A continuous state analysis function is defined to determine whether the member is in a stationary state at the data point:

[0189]

[0190] wherein N is the time window, used to determine the minimum time length of the continuous stationary state, only when the continuous N time points all satisfy the stationary state, it is determined that the member is in a stationary state;

[0191] If Status_continuous i，t = 1, it is determined that the member is in a stationary state at the data point;

[0192] If Status_continuous i，t = 0, it is determined that the member is in a moving state at the data point;

[0193] Within the determined stationary period, the refrigeration environment temperature T real，i (t) corresponding to each time point is obtained;

[0194] The average value of the temperature values within the stationary period is calculated:

[0195]

[0196] wherein M is the number of time points within the stationary period, used to calculate the average value of the temperature values within the stationary period, start is the start time of the stationary period, and end is the end time of the stationary period.

[0197] Calculate the weighted stay environment temperature:

[0198]

[0199] wherein w t is a time weight, used to weight the temperature at each time point, which can be set according to the position of the time point in the stay period or the stay duration, etc., for example, w t = t - t start + 1, to reflect the impact of temperature change on the member.

[0200] The embodiment also provides that the habit environment temperature and the habit apparent temperature of the member are analyzed, specifically:

[0201] For each indoor member, the member trajectory of the member is extracted;

[0202] For each data point on the member trajectory, the stay duration of the member at the data point is recorded, denoted as t i ;

[0203] The stay environment temperature T stay，i at the data point is obtained;

[0204] The habit environment temperature of the member is calculated:

[0205]

[0206] wherein n represents the total number of data points, i.e., the total number of data points on the member trajectory;

[0207] The outdoor environment temperature is obtained, denoted as T outdoor ;

[0208] The natural wind speed in the indoor environment is obtained, denoted as v wind ;

[0209] For each data point, the apparent temperature impact factor is calculated:

[0210] Impact_Factor i = T stay，i + k1·T outdoor + k2·v wind ;

[0211] wherein k1 and k2 are correction coefficients of the apparent temperature, used to adjust the impact of the outdoor temperature and the wind speed on the apparent temperature;

[0212] The habit apparent temperature of the member is calculated:

[0213]

[0214] Wherein, n represents the total number of data points, that is, the total number of data points on the member trajectory.

[0215] The embodiment also provides that, in combination with the current indoor and outdoor environment and member habits, indoor temperature distribution under different operation parameters of the air conditioner is simulated, specifically:

[0216] An initial temperature of the indoor space is obtained.

[0217] An outdoor temperature T outdoor is obtained.

[0218] A natural wind speed and a natural wind direction of the indoor space are obtained.

[0219] All refrigeration temperatures, air supply angles and air supply speeds of the air conditioner are obtained.

[0220] Each air supply angle of the air conditioner is mapped with each air supply speed one by one to form an air supply wind force set.

[0221] Each element in the air supply wind force set is mapped with each refrigeration temperature one by one to form a simulated air supply set, denoted as P.

[0222] For each set of air conditioner parameter combination in the simulated air supply set, the air conditioner is simulated to supply air to the indoor space.

[0223] A time point at which the air conditioner is simulated to supply air to the indoor space is set as a simulation start time.

[0224] An interval analysis duration is set.

[0225] Taking the simulation start time as a start, each interval analysis duration is taken as a time collection point.

[0226] An indoor temperature at each time collection point is calculated.

[0227]

[0228] Wherein, T real，i (i, t) is an actual indoor temperature of the i th indoor data point at time t, the indoor data point is any defined position point in the indoor space, T aircond is the air conditioner refrigeration temperature, a is a refrigeration effect attenuation coefficient, γ i is a wind speed attenuation factor of the i th data point, k te is a wind speed correction coefficient of the temperature, v wind is a natural wind speed, θ is an air supply angle, β is an outdoor environment conduction coefficient to the indoor temperature, T outdoor is an outdoor temperature.

[0229] After the air conditioner is operated for different durations under each set of air conditioner parameter combination, an indoor temperature T real，i(t) the simulated data of indoor temperature distribution stored in an array or data structure.

[0230] The embodiment further provides that the final air conditioning refrigeration temperature, air supply angle and air speed are determined, and specifically:

[0231] For each data point on the member trajectory of each member, the staying duration t of the member at the data point is recorded i ;

[0232] The weighted value of temperature change and staying time of each data point is calculated:

[0233]

[0234] Wherein, t i is the staying duration of the member at the data point i, the weighted value of temperature change and the position coincidence degree weight, reflecting the activity frequency and importance of the member at each position, T real，i (t) is the actual indoor temperature of the data point i at the time point t, T_Weighted i is the weighted value of temperature change and staying time, used for evaluating the influence of temperature change on the comfort of the member, reflecting the influence of temperature change on the member at the specific data point, combining the staying duration and the absolute value of temperature change;

[0235] The position coincidence degree weight of each data point and the member activity trajectory is calculated, and the weight reflects the relative importance of the activity frequency or staying time of the member at the data point, and is used for emphasizing the temperature influence of the frequently active area of the member:

[0236]

[0237] Wherein, t i is the staying duration of the member at the data point i, and n is the total number of data points;

[0238] The weighted temperature change of each data point is combined with the position coincidence degree weight to obtain the comprehensive temperature change T comprehensive , which combines the weighted temperature change and the position coincidence degree weight of all data points, and provides a comprehensive evaluation of the temperature change on the entire activity trajectory:

[0239]

[0240] According to the staying duration of the member at each data point, the weighted average value T simulated of the simulated temperature is calculated, reflecting the average temperature experience of the member at different positions in the room, and the weight of the staying time is considered:

[0241]

[0242] wherein T real，i represents the actual ambient temperature of the i-th data point, and t i represents the length of stay of the member at each data point, to derive a temperature value that reflects the member's sense of temperature;

[0243] a simulated temperature weighted average value T simulated_overlap that more accurately reflects the member's temperature experience in the frequently-visited area:

[0244]

[0245] wherein T real，i represents the actual ambient temperature of the i-th data point, and t i represents the length of stay of the member at each data point, to derive a temperature value that reflects the member's sense of temperature, and in combination with the position coincidence degree weight w overlap，i , a simulated temperature weighted average value is calculated that is more in line with the member's habits;

[0246] a simulated temperature weighted average value T simulated and the error Error habit between the simulated temperature weighted average value and the member's habit environment temperature T simulated , which represents the error between the simulated temperature weighted average value and the member's habit environment temperature, is used to evaluate the difference between the simulated temperature and the member's habit temperature, with a smaller error indicating that the simulation result is more in line with the member's comfort requirements:

[0247] Error simulated = |T simulated - T habit |;

[0248] a simulated temperature weighted average value T simulated_overlap that takes into account the position coincidence degree, and the error Error habit between the simulated temperature weighted average value that takes into account the position coincidence degree and the member's habit environment temperature T simulated_overlap , which represents the error between the simulated temperature weighted average value that takes into account the position coincidence degree and the member's habit environment temperature, is used to evaluate the difference between the simulated temperature that takes into account the position coincidence degree and the member's habit temperature:

[0249] Error simulated_overlap = |T simulated_overlap - T habit |;

[0250] a comprehensive error evaluation value Error_total is calculated, which is used to provide a comprehensive error evaluation index to select the air conditioning parameter combination that is most in line with the member's habit environment temperature:

[0251] Error_total = Error simulated + Errorsimulated_overlap ;

[0252] Traverse all air conditioning parameter combinations in the simulation air supply set P, record the combination with the minimum error, that is, the optimal air conditioning parameter combination most consistent with the member's habit environment temperature:

[0253]

[0254] Wherein, argmin represents finding the parameter combination in set P that makes the following expression take the minimum value, Error_total(p) is the comprehensive error evaluation value of parameter combination p;

[0255] Then the air conditioning refrigeration temperature, air supply angle and air speed in the optimal air conditioning parameter combination are the final air conditioning refrigeration temperature, air supply angle and air speed.

[0256] In this embodiment, by analyzing the habit environment temperature and the thermal sensation temperature of the member, personalized temperature regulation is provided, the comfortable experience of the member is improved, unnecessary energy consumption is reduced by optimizing the air conditioning parameters, energy utilization efficiency is improved, automatic and intelligent management of the air conditioning system is realized, manual intervention is reduced, management efficiency is improved, the system can dynamically adjust the air conditioning parameters according to the indoor and outdoor environment changes and the activity track of the member, and ensure that the indoor temperature is always in the best state.

Claims

1. An air conditioner refrigeration energy-saving self-adaptive control method based on artificial intelligence, characterized in that: The method comprises the following steps: Collecting indoor member activity trajectories; For each indoor member, extracting its activity trajectory as a member trajectory; Collecting indoor temperature data; For each member trajectory, extracting its data point set as a member trajectory point set; For each data point in the member trajectory point set, simulating the environmental temperature of each data point under different refrigeration durations to obtain a refrigeration environmental temperature; For each data point in the member trajectory point set, analyzing the stay environmental temperature of each data point; For each indoor member, analyzing the habit environmental temperature and the habit thermal sensation temperature of the member; Combining the current indoor and outdoor environments and the member habits, simulating the indoor temperature distribution under different running parameters of the air conditioner; Determining the final air conditioner refrigeration temperature, air supply angle and air speed; Controlling the air conditioner to run according to the final air conditioner refrigeration temperature, air supply angle and air speed.

2. The artificial intelligence-based air conditioner refrigeration energy-saving self-adaptive control method according to claim 1, characterized in that: The method for collecting indoor member activity trajectories comprises the following steps: Obtaining all indoor members; Defining an indoor coordinate system; Setting a fixed time interval; For each indoor member, record its location coordinates at different time points with time series to form a trajectory sequence: 3.The artificial intelligence-based air conditioner refrigeration energy-saving self-adaptive control method according to claim 1, characterized in that: Collecting indoor temperature data, specifically: Obtaining air conditioner refrigeration data when each indoor member controls the air conditioner, and forming a parameter sequence: Get all temperature collection devices in the room: J = (j1, j2,..., j n ) For each temperature acquisition device, its device location and effective acquisition range are acquired, respectively denoted as and For each indoor member, extracting its member trajectory; Setting a plurality of data points on the member trajectory to form a data point set; For each data point p l (x l , y l ) in the set of data points, the distance between it and each temperature collection device is calculated: determining an area ambient temperature for each data point in the set of data points: defining a temperature value analysis function to determine an actual ambient temperature for each data point in the set of data points: If F R_T = 1, the area ambient temperature of the data point is identified as the actual ambient temperature of the data point. If F R_T = 0, then it is presumed that the actual ambient temperature of the data point is determined.

4. The artificial intelligence-based air conditioner refrigeration energy-saving self-adaptive control method according to claim 3, characterized in that: Speculatively determining the actual environmental temperature of the data point, specifically: Obtaining the natural wind direction and natural wind speed of the data point; Calculating the temperature influence of the natural wind, set as natural wind temperature: Obtaining the regional environmental temperature of the data point; Calculating the weight coefficient of the regional environmental temperature and the natural wind temperature: Combining the regional environmental temperature and the natural wind temperature to obtain the actual environmental temperature: T estimated = a · T region + (1 - a) · T natural .

5. The artificial intelligence-based air conditioner refrigeration energy-saving self-adaptive control method according to claim 2, characterized in that: Simulating the environmental temperature of each data point under different refrigeration durations to obtain a refrigeration environmental temperature, specifically: Obtaining the actual environment temperature of the ith data point at time t-1, which is defined as the current environment temperature, and is recorded as T real，i (t-1); computing the reduction of the ambient temperature by the air conditioner at the current time instant: AT air = (T aircond - T real，i (t - 1)) · a(t); Wherein, α(t) is the attenuation coefficient of the refrigeration effect, and the specific formula is: Calculating the influence value of the natural wind on the environmental temperature at the current time: ΔT wind = v wind · cos(θ) · (T outdoor − T real，i (t−1)) · k wind ; The initial ambient temperature, air conditioning refrigeration effect and natural wind influence are added to obtain the actual ambient temperature of the ith data point at time t, which is defined as the refrigeration ambient temperature and denoted as T real，i (t): T real，i (t) = T real，i (t - 1) + ΔT air + ΔT wind .

6. The artificial intelligence-based air conditioner refrigeration energy-saving self-adaptive control method according to claim 2, characterized in that: Analyzing the stay environmental temperature of each data point, specifically: Calculate the change in position between two adjacent time points: D i，t = |x i，t+1 - x i，t | 2 + |y i，t+1 - y i，t | 2 ; A state determination function is defined to determine whether a member is in a stay state between two adjacent time points: If Status i，t = 1, then the member is determined to be in a dwell state between two adjacent time points; If Status i，t = 0, then determine that the member was in motion between the two adjacent time points; Defining a continuous state analysis function to determine whether the member is in a stay state at the data point: If Status_continuous i，t = 1, then the member is determined to be in a stationary state at this data point; If Status_continuous i，t = 0, then determine that the member is in a moving state at this data point; In a determined stay period, the refrigeration environment temperature T corresponding to each time point is acquired real，i (t); calculating an average of the temperature values over the dwell period: Calculating the weighted ambient temperature of the dwelling:

7. The artificial intelligence-based air conditioner refrigeration energy-saving self-adaptive control method according to claim 2, characterized in that: Analyzing the habit environmental temperature and the habit thermal sensation temperature of the member, specifically: For each indoor member, extracting its member trajectory; For each data point on the member trajectory, record the member's dwell duration at that data point, denoted as t i ; acquiring a dwell ambient temperature T for the data point stay，i ; calculating the habitual ambient temperature of the member: Obtain the outdoor ambient temperature, denoted as T outdoor ; Obtaining the natural wind speed in the room, denoted as v wind ; For each data point, calculating a thermal sensation influence factor: Impact_Factor i = T stay，i + k1 · T outdoor + k2 · v wind ; calculating the habitual apparent temperature for the member:

8. The artificial intelligence-based air conditioner refrigeration energy-saving self-adaptive control method according to claim 2, characterized in that: Combining the current indoor and outdoor environments and the member habits, simulating the indoor temperature distribution under different running parameters of the air conditioner, specifically: Obtaining the initial temperature of the indoor; Acquiring an outdoor temperature T outdoor ; Obtaining the natural wind speed and natural wind direction of the indoor; Obtaining all refrigeration temperatures, air supply angles and air supply speeds of the air conditioner; Mapping each air supply angle of the air conditioner with each air supply speed one by one to form an air supply wind force set; Mapping each element in the air supply wind force set with each refrigeration temperature one by one to form a simulated air supply set, denoted as P; For each set of air conditioner parameter combinations in the simulated air supply set, simulating the air supply of the air conditioner to the indoor; Defining the time point of the simulated air supply of the air conditioner to the indoor as a simulation start time; Setting an interval analysis duration; Taking the simulation start time as the starting point, forming a time collection point every interval analysis duration; Calculating the indoor temperature at each time collection point: The indoor temperature T at each location in the room at each time point after the air conditioner is operated for different lengths of time under each combination of air conditioner parameters real，i (t) storing in a data structure, obtaining simulation data of the indoor temperature distribution. 9.The artificial intelligence based air conditioner energy saving adaptive control method of claim 8, characterized in that: Determining the final air conditioner refrigeration temperature, air supply angle and air speed, specifically: For each data point on the member trajectory of each member, record the member's dwell time t at that data point i ; Calculating the weighted value of the temperature change and the stay time of each data point: Compute the position overlap weight of each data point with the member activity trajectory: Combining the weighted temperature change of each data point with the positional concordance weight gives the overall temperature change T comprehensive : According to the length of stay of the members at each data point, a weighted average value T of the simulated temperature is calculated simulated : The simulated temperature weighted average value T is calculated taking into account the degree of coincidence of the positions simulated_overlap : The weighted average of the calculated simulated temperatures T simulated The error Error habit of the member's habitual environment temperature T simulated : Error simulated = |T simulated - T habit |; The simulated temperature weighted average T considering the position coincidence degree is calculated simulated_overlap The error Error between the member habit environment temperature T habit and the simulated temperature weighted average T simulated_overlap : Error simulated_overlap = |T simulated_overlap - T habit | Calculate the total error evaluation value Error_total: Error_total = Error simulated + Error simulated_overlap ; All air conditioning parameter combinations in the simulation air supply set P are traversed, and the combination with the minimum error is recorded, that is, the optimal air conditioning parameter combination most consistent with the member's habit environment temperature: The air conditioning refrigeration temperature, the air supply angle and the air speed in the optimal air conditioning parameter combination are the final air conditioning refrigeration temperature, the air supply angle and the air speed.

Images