A method for intelligent spraying of elevator air conditioner

Abstract

The present application provides a method for intelligent spraying of elevator air conditioners, comprising: obtaining temperature and humidity sensor data in an elevator car and elevator door switch status data, including door opening time, door closing time, door opening speed, door closing speed, door opening and closing times and door opening dwell time, and calculating a first condensation water generation rate estimate based on a pre-established thermodynamic model; taking the deviation between the net condensation water generation rate and a preset condensation water generation rate safety threshold as input, combining the humidity fluctuation cycle, the daily variation law of humidity, and the coupling relationship between humidity and temperature, and adopting a fuzzy control algorithm to output the spray cycle, spray time and adjustment amount of atomization particle size, so as to realize dynamic optimization of the working parameters of the spray equipment.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to an intelligent spraying method for elevator air conditioners. Background Art

[0002] During the cooling process of the elevator air conditioner, condensed water tends to accumulate inside the air conditioner. When a certain amount of condensed water accumulates, it will overflow from the bottom of the air conditioner and drip into the elevator car, causing the interior of the car to be damp and the ground to be slippery, affecting the comfort and safety of passengers. Although this problem can be solved by installing a condensed water collection device, the capacity of the traditional collection device is limited. When the condensed water is generated at a fast speed, the collection device may not be able to be emptied in time, resulting in condensed water overflow. In order to fundamentally solve this problem, a spray device can be installed on the elevator air conditioner, and the condensed water can be atomized by the spray device to promote its evaporation and reduce the accumulation of condensed water. However, if the spray device works continuously, it will consume a lot of energy. Therefore, it is necessary to reasonably design the control strategy of the spray device to meet the condensed water treatment needs while reducing energy consumption as much as possible. To this end, a delay control function can be added to the spray device. When the air conditioner stops cooling for a period of time, the spray device is started again to avoid spraying when the air conditioner is just started, because less condensed water is generated at this time, and spraying will waste energy instead. However, how to determine the optimal delay time requires comprehensive consideration of factors such as the frequency of air conditioning use, ambient temperature and humidity, etc., which requires a large amount of data support and complex algorithm analysis. In addition, for different elevators, their operating environments and passenger capacities vary greatly, and fixed delay times are not suitable for all situations. In order to achieve the best spraying effect and energy-saving goals under different working conditions, it is necessary to dynamically adjust the delay strategy according to actual conditions. This puts higher requirements on the design of the control algorithm, and it is necessary to introduce intelligent technical means to improve the accuracy and flexibility of control. Summary of the invention

[0003] The present invention provides an elevator air conditioning intelligent spraying method, which mainly includes:

[0004] Obtain the temperature and humidity sensor data in the elevator car and the elevator door switch status data, including the door opening time, door closing time, door opening speed, door closing speed, door opening and closing times, and door opening dwell time, and calculate the estimated value of the first condensed water generation rate according to the pre-established thermodynamic model;

[0005] The liquid level sensor data of the condensate collection device is obtained, and the estimated value of the second condensate generation rate is calculated by combining the spray water return flow sensor data and the estimated value of the environmental evaporation. At the same time, the sudden change amplitude, sudden change duration, slow change amplitude, slow change duration, fluctuation frequency, fluctuation period, daily change law of the humidity in the car, and the coupling relationship data with temperature are obtained;

[0006] Determine whether the difference between the first condensate generation rate estimate and the second condensate generation rate estimate exceeds a preset threshold value. If the difference exceeds the preset threshold value, it is determined that the sensor is abnormal and an alarm signal is triggered. If the difference does not exceed the preset threshold value, a weighted average is performed based on the credibility of the two estimates and the correlation between the humidity fluctuation characteristics and the door opening and closing state to obtain a comprehensive condensate generation rate estimate.

[0007] The estimated value of the comprehensive condensate generation rate is compared with the preset condensate generation rate safety threshold. If it is greater than the safety threshold, the spray equipment is started immediately. If it is less than the safety threshold, the fuzzy control algorithm is used to dynamically adjust the delayed start time of the spray equipment according to the door opening time, door closing time, door opening speed, door closing speed, door opening and closing times, door opening dwell time, humidity mutation amplitude and duration in the car.

[0008] After the spray equipment is started, a nonlinear mapping model between nozzle pressure, atomizer frequency and condensate removal efficiency is established. The nozzle pressure and atomizer frequency are adaptively adjusted using the real-time acquired condensate collection device liquid level data and ambient temperature and humidity data to keep the spray water atomization degree at the optimal state. The spray evaporation removal rate of condensate is calculated to obtain the net condensate generation rate.

[0009] Taking the deviation between the net condensate generation rate and the preset condensate generation rate safety threshold as input, combined with the humidity fluctuation cycle, the daily variation law of humidity, and the coupling relationship between humidity and temperature, a fuzzy control algorithm is used to output the adjustment amount of the spray cycle, spray duration and atomization particle size, so as to realize the dynamic optimization of the working parameters of the spray equipment.

[0010] The technical solution provided by the embodiment of the present invention may have the following beneficial effects:

[0011] The present invention discloses a method for intelligent spraying of elevator air conditioners. The condensation water generation rate is estimated by acquiring data such as temperature and humidity in the car, door switch status, etc., combined with thermodynamic models and liquid level sensor data. At the same time, the correlation between the humidity change characteristics and the door switch status is analyzed to determine whether the sensor is abnormal. According to the comprehensive estimated condensation water generation rate, the start timing and working parameters of the spraying equipment are dynamically adjusted. The nozzle pressure and atomizer frequency are adjusted by an adaptive optimization algorithm, and the spray cycle, duration and atomization particle size are optimized by a self-learning fuzzy control algorithm to achieve precise control of condensation water. The present invention can effectively prevent the accumulation of condensation water in the elevator car, improve passenger comfort and safety, and optimize the energy efficiency of the spraying equipment, which has important practical value. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 The present invention is a flow chart of an elevator air conditioning intelligent spraying method. DETAILED DESCRIPTION

[0013] In order to further understand the content of the present invention, the present invention is described in detail in conjunction with the accompanying drawings and embodiments. The present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It is understood that the specific embodiments described herein are only used to explain the relevant inventions, rather than to limit the invention. It is also necessary to explain that, for ease of description, only the parts related to the invention are shown in the accompanying drawings.

[0014] like Figure 1 In this embodiment, a method for intelligent spraying of an elevator air conditioner may specifically include:

[0015] S101. Obtain the temperature and humidity sensor data in the elevator car and the elevator door switch status data, including door opening time, door closing time, door opening speed, door closing speed, door opening and closing times and door opening dwell time, and calculate the estimated value of the first condensation water generation rate based on a pre-established thermodynamic model.

[0016] The temperature and humidity sensor data and the elevator door switch status data in the elevator car are obtained, and the door opening time, door closing time, and door opening and closing times are calculated based on the door switch status data. The temperature and humidity data, door opening time, door closing time, and door opening and closing times are input into a pre-established thermodynamic model to obtain an estimated value of the theoretical condensation water generation rate. A regression analysis tool is used to establish a relationship model between the temperature and humidity in the elevator car and the condensation water generation rate, and the actual condensation water generation rate estimate is calculated from the data obtained in real time from the temperature and humidity sensor. The door switch status influencing factor is merged with the theoretical condensation water generation rate estimate and the actual condensation water generation rate estimate using the weighted average method to obtain a corrected first condensation water generation rate estimate.

[0017] Exemplarily, the temperature and humidity sensor data and the door switch status data in the elevator car are obtained, and the door opening time, door closing time, door opening speed, door closing speed, door opening and closing times and door opening stay time are calculated according to the door switch status data. These data are input into the pre-established thermodynamic model for calculation to obtain the estimated value of the theoretical condensation water generation rate. The thermodynamic model considers factors such as the air temperature, relative humidity, air volume, and thermal conductivity in the car, and calculates the condensation water generation rate through the energy balance equation. The relationship model between the temperature and humidity in the elevator car and the condensation water generation rate is established by using regression analysis tools, and the independent variables include the car temperature, relative humidity, door opening time, door closing time and door opening and closing times. The model is trained by historical data, and the actual condensation water generation rate is calculated from the data obtained in real time from the car temperature and humidity sensor. The air circulation condition of the car is judged according to the elevator door switch status data, and the condensation water generation rate is corrected in combination with the temperature and humidity data. If the door is in an open state, the temperature and humidity difference inside and outside the car is calculated, and the condensation water generation rate is adjusted according to the size of the temperature and humidity difference. The weighted average method is used to integrate the door switch state influencing factor with the theoretical condensate generation rate estimate and the actual condensate generation rate estimate to obtain the revised first condensate generation rate estimate. The actual condensate accumulation is measured by the electronic weighing device installed at the bottom of the elevator car, and the actual accumulation is compared with the revised condensate generation rate estimate. The recursive least squares method is used to calibrate and optimize the parameters of the thermodynamic model and the relationship model between the temperature, humidity and condensate generation rate in the elevator car in real time. The state variable is defined as the model parameter, and the observation equation is the difference between the actual measured value and the estimated value. The accuracy of the condensate generation rate estimate is improved through iterative updating. The temperature and humidity sensor data in the elevator car is obtained, such as temperature 25℃, relative humidity 60%, and the door switch state data is recorded at the same time. The calculation results show that the door opening time is 15 seconds, the door closing time is 5 seconds, the door opening speed is 0.2 m / s, the door closing speed is 0.3 m / s, the door opening and closing times are 50 times / hour, and the door opening stay time is 30 seconds. These data are input into a pre-established thermodynamic model, which takes into account factors such as air temperature, relative humidity, air volume of 3 cubic meters, and thermal conductivity of 0.024 W / m·Kelvin in the cabin. The theoretical condensation water generation rate is estimated to be 0.5 liters / hour through the energy balance equation. The multivariate linear regression tool is used to establish a relationship model between the environmental parameters in the elevator cabin and the condensation water generation rate. The independent variables include cabin temperature, relative humidity, door opening time, door closing time, and door opening and closing times. The model is trained using 1,000 sets of historical data. After obtaining the regression coefficient, the actual condensation water generation rate is estimated to be 0.6 liters / hour from the sensor data obtained in real time. The cabin air circulation is judged according to the elevator door switch status, and the condensation water generation rate is corrected in combination with the temperature and humidity data. When the door is in the open state, the temperature difference between the inside and outside of the cabin is calculated to be 5°C and the humidity difference is 20%. The condensation water generation rate is adjusted using the preset influence factor of 0.8.By weighted average method, the theoretical condensate generation rate estimate and the actual condensate generation rate estimate are combined with weights of 0.6 and 0.4, respectively, to obtain a revised condensate generation rate estimate of 0.55 liters / hour. A high-precision electronic weighing device with an accuracy of 0.01 kg is installed at the bottom of the elevator car to measure the actual accumulation of condensate. The actual accumulation is recorded once an hour and compared with the revised condensate generation rate estimate. Using the recursive least squares method, the initial parameter estimates and covariance matrix are set, the state variables are defined as model parameters, and the observation equation is the difference between the actual measured value and the estimated value. Through iterative updating, the gain matrix is ​​calculated, the parameter estimates and the covariance matrix are updated, and finally the optimized model parameters are obtained, which improves the accuracy of the condensate generation rate estimation to an error range of ±0.05 liters / hour.

[0018] S102, obtain the liquid level sensor data of the condensate collection device, and calculate the estimated value of the second condensate generation rate by combining the spray water return flow sensor data and the estimated value of the environmental evaporation. At the same time, obtain the sudden change amplitude, sudden change duration, slow change amplitude, slow change duration, fluctuation frequency, fluctuation period, daily change law of the humidity in the car, and the coupling relationship data with temperature.

[0019] The liquid level sensor data of the condensate collection device is obtained, and the spray water return flow sensor data is recorded at the same time; the second condensate generation rate estimate is calculated based on the liquid level sensor data and the spray water return flow sensor data. The humidity sensor data in the car is processed to calculate the humidity change rate; the humidity mutation amplitude, mutation duration, slow change amplitude, and slow change duration are determined based on the humidity change rate and the preset threshold. The spectrum of the humidity data is calculated using fast Fourier transform; the fluctuation frequency and fluctuation period of the humidity data are extracted from the spectrum. The Pearson correlation coefficient of the temperature data and the humidity data in the car is calculated; the temperature and humidity coupling relationship data is obtained based on the Pearson correlation coefficient. A multivariate nonlinear regression equation is established, wherein the independent variables of the multivariate nonlinear regression equation include the humidity mutation amplitude, the mutation duration, the slow change amplitude, the slow change duration, the fluctuation frequency, the fluctuation period, and the temperature and humidity coupling relationship data, and the dependent variable is the second condensate generation rate estimate; a condensate generation rate prediction model is constructed based on the multivariate nonlinear regression equation.

[0020] Exemplarily, the liquid level sensor data of the condensate collection device is obtained, and the spray water return sensor data is recorded at the same time. The environmental evaporation estimation formula is obtained based on the regression analysis of historical data, and the environmental evaporation estimation value is calculated. The second condensate generation rate estimation value is calculated by the difference between the liquid level change, the spray water return and the environmental evaporation. The humidity sensor data in the car is processed by the time series analysis tool, the humidity change rate is calculated by the difference method, and the preset threshold is set to judge the sudden change and slow change, so as to obtain the sudden change amplitude, sudden change duration, slow change amplitude and slow change duration of the humidity. The spectrum of the humidity data is calculated by fast Fourier transform, and the fluctuation frequency and fluctuation period are extracted. The humidity data is smoothed by the moving average method to extract the daily change law. The correlation between the temperature and humidity data in the car is calculated by the Pearson correlation coefficient, and the temperature and humidity coupling relationship data is obtained. Combined with the second condensate generation rate estimation value, humidity change characteristics and temperature and humidity coupling relationship, a multivariate nonlinear regression equation is established to construct a condensate generation rate prediction model. The independent variables include humidity mutation amplitude, duration of mutation, slow change amplitude, duration of slow change, fluctuation frequency, fluctuation period, intraday change amplitude and temperature-humidity coupling coefficient, and the dependent variable is the condensate generation rate. The sliding time window method is used to monitor the humidity data in the car in real time. If the humidity mutation exceeds the preset threshold, the alarm mechanism is triggered. According to the output results of the prediction model, the working parameters of the condensate collection device are dynamically adjusted, such as increasing or decreasing the collection frequency and adjusting the collection container capacity. The dynamic optimization of the condensate collection process is realized to improve the collection efficiency.

[0021] The liquid level sensor of the condensate collection device collects data every 5 minutes and records the change in liquid level height. At the same time, the spray water return flow sensor records the return water volume. Based on the historical data of the past 30 days, the environmental evaporation estimation formula is obtained using polynomial regression: E = 0.02T^2 + 0.5H-0.1, where E is the evaporation volume (ml / hour), T is the temperature (℃), and H is the relative humidity (%). The current environmental evaporation estimate is calculated to be 50 ml / hour. The second condensate generation rate is estimated to be 50 ml / hour by the difference between the liquid level change of 300 ml / hour, the spray water return flow of 200 ml / hour and the environmental evaporation of 50 ml / hour. The humidity sensor in the car collects data once a second, and the humidity change rate is calculated using the difference method of a 10-second time window. 5% / second is set as the mutation threshold and 1% / second is set as the slow change threshold. It is judged that the humidity mutation amplitude is 8%, the mutation duration is 3 seconds, the slow change amplitude is 2%, and the slow change duration is 30 seconds. The humidity data spectrum was calculated using a 1024-point fast Fourier transform, and the main fluctuation frequency of 0.05 Hz and the fluctuation period of 20 seconds were extracted. A 60-minute moving average window was used to smooth the humidity data and extract the intraday variation pattern. It was found that the humidity had two peaks at 9:00-11:00 and 14:00-16:00. The Pearson correlation coefficient of the temperature and humidity data was calculated to be 0.75, indicating a strong positive correlation. A multivariate nonlinear regression equation was established, with the independent variables including the above 8 features and the dependent variable being the condensate generation rate. A 5-minute sliding time window was used to monitor the cabin humidity in real time. If a sudden change in humidity exceeding the 8% / second threshold was detected, an alarm was triggered. According to the output results of the prediction model, when the predicted condensate generation rate exceeded 100 ml / hour, the collection frequency was adjusted from once per hour to once every 30 minutes, and the collection container capacity was increased from 2 liters to 5 liters to achieve dynamic optimization of the condensate collection process.

[0022] S103, determine whether the difference between the first condensation water generation rate estimate and the second condensation water generation rate estimate exceeds a preset threshold value. If so, determine that the sensor is abnormal and trigger an alarm signal. If not, perform a weighted average based on the credibility of the two estimates and the correlation between the humidity fluctuation characteristics and the door opening and closing status to obtain a comprehensive condensation water generation rate estimate.

[0023] Obtain the first condensate generation rate estimate and the second condensate generation rate estimate, and calculate the difference between the first condensate generation rate estimate and the second condensate generation rate estimate; determine whether the difference exceeds the difference threshold according to a preset difference threshold, and if the difference exceeds the difference threshold, determine that there is a sensor abnormality. Calculate the credibility index of the first condensate generation rate estimate and the second condensate generation rate estimate based on historical data statistical analysis; use the mean absolute error and standard deviation to evaluate the accuracy of the first condensate generation rate estimate and the second condensate generation rate estimate, and obtain the weight coefficient of the first condensate generation rate estimate and the second condensate generation rate estimate. Use time series tools to process humidity fluctuation data, extract the fluctuation amplitude, period and trend characteristics of the humidity fluctuation data; obtain elevator door switch status data, and calculate the correlation coefficient between the characteristics of the humidity fluctuation data and the elevator door switch status data through the Pearson correlation coefficient. The weighted average method is adopted to perform a comprehensive calculation based on the first condensation water generation rate estimate, the second condensation water generation rate estimate, the characteristics of the humidity fluctuation data and the correlation coefficient; an adjustment coefficient is set, and the weight coefficient is dynamically adjusted according to the fluctuation amplitude of the humidity fluctuation data and the frequency of the elevator door switch status data to obtain a comprehensive condensation water generation rate estimate.

[0024] Exemplarily, the difference between the first condensate generation rate estimate and the second condensate generation rate estimate is calculated, and whether there is a sensor abnormality is determined by a preset difference threshold. If the difference exceeds the preset threshold, an alarm signal is triggered, the abnormal situation is recorded, and the backup sensor is enabled. The abnormal data is marked for subsequent analysis and processing. According to the statistical analysis of historical data, the credibility index of the first condensate generation rate estimate and the second condensate generation rate estimate is calculated. The accuracy of the two estimation methods is evaluated by the mean absolute error and the standard deviation, and the respective weight coefficients are obtained. The humidity fluctuation data is processed by the time series tool to extract the characteristics of the fluctuation amplitude, cycle and trend. At the same time, the elevator door switch state data is obtained, and the correlation coefficient between the humidity fluctuation characteristics and the switch door state is calculated by the Pearson correlation coefficient. The weighted average method is used to comprehensively calculate the correlation between the first condensate generation rate estimate, the second condensate generation rate estimate, the humidity fluctuation characteristics and the switch door state. The adjustment coefficient is set, and the weight is dynamically adjusted according to the humidity fluctuation amplitude and the switch door frequency to obtain the final comprehensive condensate generation rate estimate. The sliding time window method is used to regularly update the preset threshold and weight coefficient. According to the recent operation data, the difference threshold and the accuracy of the estimation method are recalculated, and the parameters are dynamically adjusted to adapt to different operating environments and conditions. The elevator condensate monitoring system collects data every 5 minutes, and the first condensate generation rate is estimated to be 0.5 liters / hour, the second condensate generation rate is estimated to be 0.6 liters / hour, and the difference is 0.1 liters / hour. The preset difference threshold is 0.2 liters / hour, and the current difference does not exceed the threshold, so the sensor is judged to be normal. Based on the historical data of the past 30 days, the average absolute error of the first estimation method is calculated to be 0.05 liters / hour, and the standard deviation is 0.03 liters / hour; the average absolute error of the second estimation method is 0.04 liters / hour, and the standard deviation is 0.02 liters / hour. Based on this calculation, the weight of the first estimation method is 0.45, and the weight of the second estimation method is 0.55. The humidity fluctuation data is analyzed using fast Fourier transform, and the main fluctuation frequency is extracted as 0.05 Hz, and the fluctuation amplitude is 5% relative humidity. The elevator door switch status data shows that the door is opened and closed 30 times per hour on average. The Pearson correlation coefficient between humidity fluctuation and door opening and closing status is calculated to be 0.75, indicating a strong correlation. The system sets the humidity fluctuation adjustment coefficient to 1.1 and the door opening and closing frequency adjustment coefficient to 0.9. The two estimated values, humidity fluctuation characteristics, and door opening and closing status are weighted averaged to obtain an estimated comprehensive condensate generation rate of 0.57 liters / hour. Using a 24-hour sliding time window, the preset threshold and weight coefficient are updated every 6 hours. Based on the operating data of the last 24 hours, the difference threshold is recalculated to 0.18 liters / hour, the new weight of the first estimation method is 0.47, and the new weight of the second estimation method is 0.53, which dynamically adapts to the current operating environment.

[0025] S104. Compare the estimated value of the comprehensive condensate generation rate with the preset condensate generation rate safety threshold. If it is greater than the safety threshold, start the spray equipment immediately. If it is less than the safety threshold, use a fuzzy control algorithm to dynamically adjust the delayed start time of the spray equipment based on the door opening time, door closing time, door opening speed, door closing speed, door opening and closing times, door opening dwell time, and humidity mutation amplitude and duration in the car.

[0026] Obtain an estimated value of the comprehensive condensate generation rate and compare it with a preset condensate generation rate safety threshold; if the estimated value of the comprehensive condensate generation rate is greater than the safety threshold, send a start signal to the spray equipment controller; if the estimated value of the comprehensive condensate generation rate is less than the safety threshold, obtain elevator operation parameters, which include door opening time, door closing time, door opening speed, door closing speed, door opening and closing times, and door opening dwell time; perform quantization and standardization processing based on the elevator operation parameters to obtain standardized elevator operation parameters; input the standardized elevator operation parameters and the humidity mutation data in the car into a preset fuzzy controller, and calculate the delayed start time, operation time and spray intensity of the spray equipment through fuzzification, reasoning and defuzzification processes; set the timer and control parameters of the spray equipment according to the result output by the fuzzy controller, automatically trigger the start of the spray equipment after a specified time, and control the operation time and intensity of the spray equipment.

[0027] Exemplarily, the estimated value of the comprehensive condensate generation rate is compared with the preset condensate generation rate safety threshold. If it is greater than the safety threshold, a start signal is immediately sent to the spray equipment controller to trigger the immediate operation of the spray equipment. If the estimated value of the comprehensive condensate generation rate is less than the safety threshold, the elevator operation parameters are obtained, including the door opening time, door closing time, door opening speed, door closing speed, door opening and closing times, and door opening dwell time, and the humidity mutation amplitude and mutation duration data in the car are recorded at the same time. The acquired parameters are quantified and standardized, the time parameters are converted into percentage form, the speed parameters are normalized to the range of 0-1, and the door opening and closing times are converted to hourly frequency. The standardized elevator operation parameters and humidity data are input into the pre-designed fuzzy controller, and the delayed start time, operation time and spray intensity of the spray equipment are calculated through the fuzzification, reasoning and defuzzification process. The design of the fuzzy controller includes the definition of the membership function of the input variables, the establishment of the fuzzy rule base and the application of the Mamdan i reasoning method. According to the output of the fuzzy controller, the timer and control parameters of the spray equipment are set to automatically trigger the start of the spray equipment after the specified time, and control its running time and intensity to achieve dynamic control of condensate generation. According to the condensate generation after spraying, the parameters and rules of the fuzzy controller are dynamically adjusted using the feedback mechanism to optimize the control effect. For possible abnormal situations, such as sensor failure or unresponsive spray equipment, a backup control scheme is set to ensure the reliable operation of the system. The elevator condensate control system monitors the comprehensive condensate generation rate in real time. The current estimated value is 0.8 liters / hour, and the preset safety threshold is 1.0 liters / hour. Since the estimated value is less than the safety threshold, the system begins to collect elevator operation parameters including door opening time of 15 seconds (normalized to 25%), door closing time of 5 seconds (8.3%), door opening speed of 0.5 m / s (normalized to 0.8), door closing speed of 0.6 m / s (0.9), door opening and closing times of 50 times / hour, and door opening dwell time of 30 seconds (50%). At the same time, the humidity mutation amplitude of 5% and the duration of the mutation of 3 seconds in the car are recorded. The fuzzy controller uses a triangular membership function, and the input variables are divided into three levels: "low", "medium" and "high". The output variables are delayed start time 0-30 minutes, running time 5-20 minutes and spray intensity 30%-100%. Mamdan i reasoning method is applied to perform fuzzy reasoning based on 25 IF-THEN rules. Defuzzification uses the centroid method to obtain a delayed start time of 10 minutes, a running time of 15 minutes, and a spray intensity of 60%. The system sets the timer and control parameters of the spray equipment accordingly. After the spraying is completed, the sensor detects that the condensation water generation rate has dropped to 0.5 liters / hour. The feedback mechanism inputs this data into the optimization algorithm to adjust the fuzzy rule weights and improve the control accuracy under low humidity conditions.During operation, if the humidity sensor has no data output for 5 minutes, it will automatically switch to the fixed cycle control mode based on historical data, starting the spraying for 10 minutes every 2 hours with an intensity of 50% until the sensor returns to normal.

[0028] S105. After the spray equipment is started, a nonlinear mapping model between nozzle pressure, atomizer frequency and condensate removal efficiency is established. By using the real-time acquired liquid level data of the condensate collection device and the ambient temperature and humidity data, the nozzle pressure and atomizer frequency are adaptively adjusted to keep the spray water atomization degree continuously at the optimal state. The spray evaporation removal rate of condensate is calculated to obtain the net condensate generation rate.

[0029] Support vector regression is used to train the collected historical data to obtain a nonlinear mapping model between nozzle pressure, atomizer frequency and condensate removal efficiency; the liquid level data of the condensate collection device and the ambient temperature and humidity data are obtained, and the current condensate removal efficiency is calculated according to the nonlinear mapping model; if the condensate removal efficiency is lower than the preset threshold, the nozzle pressure and the atomizer frequency are optimized and adjusted using the gradient descent method; the spray evaporation removal rate of condensate is calculated based on the optimized nozzle pressure and atomizer frequency, combined with the liquid level change rate, the spray water volume and the predetermined evaporation coefficient; the net condensate generation rate is obtained by subtracting the estimated value of the comprehensive condensate generation rate from the spray evaporation removal rate of condensate

[0030] Exemplarily, based on historical data, a nonlinear mapping model between nozzle pressure, atomizer frequency and condensate removal efficiency is established, and support vector regression is used to train the collected data to obtain the initial model parameters. The liquid level data and environmental temperature and humidity data of the condensate collection device are obtained in real time, and the condensate removal efficiency is compensated by the preset temperature and humidity correction coefficient. These data are input into the nonlinear mapping model to calculate the current condensate removal efficiency. Using the gradient descent method as an adaptive optimization tool, with the condensate removal efficiency as the objective function, the nozzle pressure and atomizer frequency are optimized and adjusted in real time to make the spray water atomization degree reach the predefined optimal state. The optimal state is determined by setting the condensate removal efficiency threshold and energy consumption index. When the condensate removal efficiency exceeds 90% and the energy consumption is lower than the preset value, it is determined to be the optimal state. According to the optimized nozzle pressure and atomizer frequency, combined with the liquid level change rate, the spray water volume and the predetermined evaporation coefficient, the spray evaporation removal rate of condensate is calculated. Finally, the spray evaporation removal rate of condensate is subtracted from the estimated value of the comprehensive condensate generation rate to obtain the net condensate generation rate as a quantitative indicator of the system operation effect. The elevator condensate control system first collected 1000 sets of historical data, including nozzle pressure 0.5-2.5MPa, atomizer frequency 20-80Hz and corresponding condensate removal efficiency 60%-95%. Using the support vector regression algorithm, the kernel function selected RBF, the penalty factor C was set to 100, and the eps il on was set to 0.1, and the nonlinear mapping model was trained. Real-time monitoring shows that the current condensate collection device liquid level is 80%, the ambient temperature is 25℃, and the relative humidity is 60%. The system applies the preset temperature and humidity correction coefficient of 1.05 to compensate for the condensate removal efficiency. These data are input into the model, and the current condensate removal efficiency is calculated to be 85%. The gradient descent optimization algorithm is started, the learning rate is set to 0.01, the number of iterations is 50, and the nozzle pressure and atomizer frequency are adjusted with the condensate removal efficiency as the objective function. The optimization results show that when the nozzle pressure is adjusted to 1.8MPa and the atomizer frequency is adjusted to 65Hz, the condensate removal efficiency is increased to 92%, and the energy consumption index is 0.8kWh / m 3 , below the preset threshold of 1.0kWh / m 3 , reaching the best state. According to the optimized parameters, combined with the liquid level change rate of 0.5% / min, the spray water volume of 100L / h and the predetermined evaporation coefficient of 0.8, the spray evaporation removal rate of condensed water is calculated to be 2.5L / h. This rate is subtracted from the previously estimated comprehensive condensed water generation rate of 3.0L / h to obtain a net condensed water generation rate of 0.5L / h, which is used as a quantitative indicator of the current operation effect of the system.

[0031] S106. Taking the deviation between the net condensate generation rate and the preset condensate generation rate safety threshold as input, combined with the humidity fluctuation cycle, the daily variation law of humidity, and the coupling relationship between humidity and temperature, a fuzzy control algorithm is used to output the spray cycle, spray duration, and the adjustment amount of atomized particle size, so as to realize dynamic optimization of the working parameters of the spray equipment.

[0032] The net condensate generation rate is obtained, and the deviation value between the net condensate generation rate and the preset condensate generation rate safety threshold is calculated; according to the deviation value, the humidity data is processed by the fast Fourier transform method to obtain the humidity fluctuation period and the daily variation law; the deviation value, the humidity fluctuation period, the daily variation law and the temperature and humidity coupling relationship are input into the self-learning fuzzy controller, and the self-learning fuzzy controller includes a triangular membership function and a trapezoidal membership function; the recursive least squares method is applied to the self-learning fuzzy controller to perform online learning and parameter adjustment to control the weights and membership function parameters of the error optimization fuzzy rules; after the self-learning fuzzy controller outputs the control amount, the control amount is converted into the adjustment amount of the spray cycle, spray duration and atomization particle size; the actuator receives the adjustment amount, and adjusts the spray equipment in real time according to the adjustment amount to achieve dynamic optimization control.

[0033] Exemplarily, the deviation between the net condensate generation rate and the preset condensate generation rate safety threshold is calculated, the humidity fluctuation period and daily variation law are obtained by fast Fourier transform, and the coupling relationship between humidity and temperature is obtained by mutual information entropy and conditional probability distribution analysis. A self-learning fuzzy controller is constructed, and the deviation value, humidity fluctuation period, daily variation law and temperature-humidity coupling relationship are used as input variables, triangular and trapezoidal membership functions are designed, and the initial fuzzy rule base is pre-formulated according to historical data. The recursive least squares method is used to perform online learning and parameter adjustment on the fuzzy controller, and the weights and membership function parameters of the fuzzy rules are continuously optimized by minimizing the control error. According to the output result of the self-learning fuzzy controller, the mapping relationship between the control quantity and the actual equipment parameters is established, and the controller output is converted into the specific adjustment amount of the spray cycle, spray duration and atomized particle size through a lookup table or linear transformation. The actuator adjusts the spray equipment in real time according to the converted parameters to achieve dynamic optimization control. By real-time monitoring of the control effect, the root mean square value of the control error is calculated. When the error exceeds the preset threshold, the forgetting factor and step size parameters of the learning algorithm are automatically adjusted to improve the system's adaptive ability and control accuracy. The elevator condensate control system monitors the net condensate generation rate of 0.8 liters / hour in real time, and the preset safety threshold is 1.0 liters / hour. The calculated deviation value is -0.2 liters / hour. The humidity data is analyzed by 1024-point fast Fourier transform, and the main fluctuation period is 4 hours. The daily variation pattern shows that 9:00-11:00 and 14:00-16:00 are the peak periods. The temperature and humidity coupling coefficient is 0.75 obtained by calculating the mutual information entropy, indicating a strong correlation. The self-learning fuzzy controller designs the membership function of 5 input variables, and each variable is divided into 3 levels. For example, the membership function of the deviation value is divided into "negative large", "close to zero" and "positive large". The initial fuzzy rule base contains 243 rules, such as "if the deviation is negative and the humidity fluctuates frequently, increase the spraying frequency". The initial forgetting factor of the recursive least squares method is set to 0.98, and the fuzzy rule weights and membership function parameters are updated at each iteration. The controller output range is [-1,1]. The output is mapped to specific adjustment amounts through a lookup table, including a spray cycle adjustment range of ±30 minutes, a spray duration adjustment range of ±5 minutes, and an atomized particle size adjustment range of ±10 microns. Based on the mapping results, the actuator adjusts the spray cycle from 120 minutes to 100 minutes, increases the spray duration from 10 minutes to 13 minutes, and reduces the atomized particle size from 50 microns to 45 microns. The system calculates the root mean square value of the control error once an hour. The current value is 0.15 liters / hour, which exceeds the preset threshold of 0.1 liters / hour. The forgetting factor is automatically adjusted to 0.95 to speed up learning and improve control accuracy.

[0034] The above embodiment is only one of the preferred implementation modes of the present invention and should not be used to limit the protection scope of the present invention. Any changes or modifications that are made to the main design concept and spirit of the present invention and have no substantive significance, and the technical problems they solve are still consistent with the present invention, should be included in the protection scope of the present invention.

Claims

1. A method for intelligent spraying of elevator air conditioner, characterized in that: The method comprises: Obtain the temperature and humidity sensor data in the elevator car and the elevator door switch status data, including the door opening time, door closing time, door opening speed, door closing speed, door opening and closing times, and door opening dwell time, and calculate the estimated value of the first condensed water generation rate according to the pre-established thermodynamic model; The liquid level sensor data of the condensate collection device is obtained, and the estimated value of the second condensate generation rate is calculated by combining the spray water return flow sensor data and the estimated value of the environmental evaporation. At the same time, the sudden change amplitude, sudden change duration, slow change amplitude, slow change duration, fluctuation frequency, fluctuation period, daily change law of the humidity in the car, and the coupling relationship data with temperature are obtained; Determine whether the difference between the first condensate generation rate estimate and the second condensate generation rate estimate exceeds a preset threshold value. If the difference exceeds the preset threshold value, it is determined that the sensor is abnormal and an alarm signal is triggered. If the difference does not exceed the preset threshold value, a weighted average is performed based on the credibility of the two estimates and the correlation between the humidity fluctuation characteristics and the door opening and closing state to obtain a comprehensive condensate generation rate estimate. The estimated value of the comprehensive condensate generation rate is compared with the preset condensate generation rate safety threshold. If it is greater than the safety threshold, the spray equipment is started immediately. If it is less than the safety threshold, the fuzzy control algorithm is used to dynamically adjust the delayed start time of the spray equipment according to the door opening time, door closing time, door opening speed, door closing speed, door opening and closing times, door opening dwell time, humidity mutation amplitude and duration in the car. After the spray equipment is started, a nonlinear mapping model between nozzle pressure, atomizer frequency and condensate removal efficiency is established. The nozzle pressure and atomizer frequency are adaptively adjusted using the real-time acquired condensate collection device liquid level data and ambient temperature and humidity data to keep the spray water atomization degree at the optimal state. The spray evaporation removal rate of condensate is calculated to obtain the net condensate generation rate. Taking the deviation between the net condensate generation rate and the preset condensate generation rate safety threshold as input, combined with the humidity fluctuation cycle, the daily variation law of humidity, and the coupling relationship between humidity and temperature, a fuzzy control algorithm is used to output the adjustment amount of the spray cycle, spray duration and atomization particle size, so as to realize the dynamic optimization of the working parameters of the spray equipment.

2. The method according to claim 1, characterized in that The method obtains the temperature and humidity sensor data in the elevator car and the elevator door switch state data, including the door opening time, door closing time, door opening speed, door closing speed, door opening and closing times, and door opening dwell time, and calculates the first condensation water generation rate estimate value according to the pre-established thermodynamic model, including: Obtain the temperature and humidity sensor data in the elevator car and the elevator door switch status data, and calculate the door opening time, door closing time and door opening and closing times according to the door switch status data; The temperature and humidity data, door opening time, door closing time and door opening and closing times are input into a pre-established thermodynamic model to obtain an estimated value of the theoretical condensation water generation rate; A regression analysis tool is used to establish a relationship model between the temperature and humidity in the elevator car and the condensation water generation rate, and the actual condensation water generation rate estimate is calculated from the real-time data obtained by the temperature and humidity sensors; The door switch state influencing factor is integrated with the theoretical condensate generation rate estimate and the actual condensate generation rate estimate using the weighted average method to obtain a corrected first condensate generation rate estimate.

3. The method according to claim 1, characterized in that The liquid level sensor data of the condensate collection device is obtained, and the second condensate generation rate estimate is calculated in combination with the spray water return flow sensor data and the environmental evaporation estimate. At the same time, the sudden change amplitude, sudden change duration, slow change amplitude, slow change duration, fluctuation frequency, fluctuation period, daily change law of humidity in the car, and coupling relationship data with temperature are obtained, including: Obtain the liquid level sensor data of the condensate collection device and record the spray water return flow sensor data; Calculate the estimated value of the second condensed water generation rate according to the liquid level sensor data and the spray water return flow sensor data; Process the humidity sensor data in the car and calculate the humidity change rate; Determine the humidity mutation amplitude, mutation duration, slow change amplitude and slow change duration according to the humidity change rate and the preset threshold; Calculate the frequency spectrum of humidity data using Fast Fourier Transform; extracting the fluctuation frequency and fluctuation period of the humidity data from the frequency spectrum; Calculate the Pearson correlation coefficient of the temperature data and humidity data in the car; The temperature and humidity coupling relationship data were obtained based on the Pearson correlation coefficient; Establishing a multivariate nonlinear regression equation, wherein the independent variables of the multivariate nonlinear regression equation include the humidity mutation amplitude, the mutation duration, the slow change amplitude, the slow change duration, the fluctuation frequency, the fluctuation period and the temperature and humidity coupling relationship data, and the dependent variable is the second condensed water generation rate estimation value; A condensate generation rate prediction model is constructed based on the multivariate nonlinear regression equation.

4. The method according to claim 3, characterized in that The method of determining whether the difference between the first condensation water generation rate estimation value and the second condensation water generation rate estimation value exceeds a preset threshold value, if it exceeds the preset threshold value, determining that the sensor is abnormal and triggering an alarm signal, if it does not exceed the preset threshold value, performing weighted averaging based on the credibility of the two estimation values, combined with the correlation between the humidity fluctuation characteristics and the door opening and closing state, to obtain a comprehensive condensation water generation rate estimation value, includes: Obtaining a first condensed water generation rate estimation value and a second condensed water generation rate estimation value, and calculating a difference between the first condensed water generation rate estimation value and the second condensed water generation rate estimation value; Determining whether the difference value exceeds the difference threshold according to a preset difference threshold, and if the difference value exceeds the difference threshold, determining that a sensor abnormality exists; Calculating the credibility index of the first condensed water generation rate estimate value and the second condensed water generation rate estimate value based on historical data statistical analysis; Using mean absolute error and standard deviation to evaluate the accuracy of the first condensed water generation rate estimate and the second condensed water generation rate estimate, and obtaining weight coefficients of the first condensed water generation rate estimate and the second condensed water generation rate estimate; Using time series tools to process humidity fluctuation data, extracting fluctuation amplitude, cycle and trend characteristics of the humidity fluctuation data; Acquire elevator door switch state data, and calculate the correlation coefficient between the characteristics of the humidity fluctuation data and the elevator door switch state data by using the Pearson correlation coefficient; Using a weighted average method, a comprehensive calculation is performed based on the first condensation water generation rate estimate value, the second condensation water generation rate estimate value, the characteristics of the humidity fluctuation data, and the correlation coefficient; An adjustment coefficient is set, and the weight coefficient is dynamically adjusted according to the fluctuation amplitude of the humidity fluctuation data and the frequency of the elevator door switch state data to obtain an estimated value of the comprehensive condensation water generation rate.

5. The method according to claim 1, characterized in that The estimated value of the comprehensive condensate generation rate is compared with the preset condensate generation rate safety threshold. If it is greater than the safety threshold, the spray equipment is started immediately. If it is less than the safety threshold, the delayed start time of the spray equipment is dynamically adjusted using a fuzzy control algorithm according to the door opening time, door closing time, door opening speed, door closing speed, door opening and closing times, door opening dwell time, and humidity mutation amplitude and duration in the car, including: Obtain an estimated value of the comprehensive condensate generation rate and compare it with a preset condensate generation rate safety threshold; If the estimated value of the comprehensive condensate generation rate is greater than the safety threshold, sending a start signal to the spray equipment controller; If the estimated value of the comprehensive condensate generation rate is less than the safety threshold, the elevator operation parameters are obtained, wherein the elevator operation parameters include door opening time, door closing time, door opening speed, door closing speed, door opening and closing times, and door opening dwell time; Quantifying and standardizing the elevator operating parameters to obtain standardized elevator operating parameters; The standardized elevator operation parameters and the humidity mutation data in the car are input into a preset fuzzy controller, and the delayed start time, operation duration and spray intensity of the spray equipment are calculated through fuzzification, reasoning and defuzzification processes; According to the output of the fuzzy controller, the timer and control parameters of the spray equipment are set to automatically trigger the start of the spray equipment after the specified time, and control the running time and intensity of the spray equipment.

6. The method according to claim 1, characterized in that After the spraying equipment is started, a nonlinear mapping model between nozzle pressure, atomizer frequency and condensed water removal efficiency is established, and the nozzle pressure and atomizer frequency are adaptively adjusted by using the real-time acquired condensed water collection device liquid level data and ambient temperature and humidity data, so that the spray water atomization degree is continuously maintained at the optimal state, and the spray evaporation removal rate of condensed water is calculated to obtain the net condensed water generation rate, including: Support vector regression was used to train the collected historical data to obtain a nonlinear mapping model between nozzle pressure and atomizer frequency and condensate removal efficiency; Acquire the liquid level data of the condensate collection device and the ambient temperature and humidity data, and calculate the current condensate removal efficiency according to the nonlinear mapping model; If the condensate removal efficiency is lower than a preset threshold, the nozzle pressure and the atomizer frequency are optimized and adjusted using a gradient descent method; The spray evaporation removal rate of condensed water is calculated based on the optimized nozzle pressure and atomizer frequency, combined with the liquid level change rate, spray water volume and the predetermined evaporation coefficient; The estimated value of the comprehensive condensate generation rate is subtracted from the spray evaporation condensate removal rate to obtain the net condensate generation rate.

7. The method according to claim 1, characterized in that The method uses the deviation between the net condensate generation rate and the preset condensate generation rate safety threshold as input, combines the humidity fluctuation cycle, the daily change law of humidity, and the coupling relationship between humidity and temperature, and adopts a fuzzy control algorithm to output the spray cycle, spray duration, and the adjustment amount of atomized particle size to achieve dynamic optimization of the working parameters of the spray equipment, including: Obtaining a net condensed water generation rate, and calculating a deviation value between the net condensed water generation rate and a preset condensed water generation rate safety threshold; According to the deviation value, the humidity data is processed by using a fast Fourier transform method to obtain the fluctuation period and daily variation law of humidity; Inputting the deviation value, the humidity fluctuation period, the intra-day variation law and the temperature-humidity coupling relationship into a self-learning fuzzy controller, wherein the self-learning fuzzy controller includes a triangular membership function and a trapezoidal membership function; Applying recursive least squares method to the self-learning fuzzy controller, performing online learning and parameter adjustment, and controlling the weights and membership function parameters of the error-optimized fuzzy rules; After the self-learning fuzzy controller outputs the control amount, the control amount is converted into the adjustment amount of the spray cycle, the spray duration and the atomized particle size; The actuator receives the adjustment amount and adjusts the spraying equipment in real time according to the adjustment amount to achieve dynamic optimization control.

Images