Intelligent integrated kitchen data collection method and system based on internet of things

Abstract

The application relates to the technical field of data processing, and discloses an intelligent integrated kitchen data acquisition method and system based on an Internet of Things, which comprises the following steps: acquiring smoke concentration data in a kitchen in real time, correcting the real-time smoke concentration data, uploading the corrected smoke concentration data to a cloud platform, and the correction process comprises the following steps: acquiring a historical smoke concentration time sequence and a historical environment characteristic time sequence, constructing a target sequence of the historical smoke concentration time sequence and the historical environment characteristic time sequence, calculating static correlation and dynamic correlation between the environment characteristics and the smoke concentration, calculating a compensation value of the smoke concentration data based on the static correlation and the dynamic correlation, and obtaining corrected smoke concentration data by combining the compensation value and the real-time smoke concentration data. The compensation value of different environment characteristics to the smoke concentration and the static correlation and the dynamic correlation are fused and compensated, so that the accuracy of the smoke concentration data uploaded to the cloud platform is improved.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a smart integrated kitchen data acquisition method and system based on the Internet of Things. Background Technology

[0002] In modern smart kitchen systems, sensors in the kitchen environment can be used not only to monitor factors such as temperature, humidity, and gas concentration, but also to achieve fire early warning and air quality management by monitoring smoke concentration. However, smoke sensors can be affected by factors such as cooking fumes, water vapor, temperature changes, and airflow, leading to false alarms or missed alarms. To improve the accuracy of smoke sensors, existing technologies use data correction and compensation from temperature sensors to adjust the output of optical smoke sensors.

[0003] For example, Chinese patent application CN118447640A discloses a method, system, and device for detecting inhaled smoke alarms based on temperature compensation. The method includes: calculating the initial zero point of an ionization smoke sensor according to a first strategy; processing the detection data using a mean clustering algorithm to obtain a variance value to determine whether the current zero point is stable; calculating the dynamic zero point of the ionization smoke sensor according to a second strategy; performing temperature compensation calculation on the initial zero point according to a pre-established zero point temperature compensation matrix to obtain an initial compensated zero point; performing linear fitting; calculating the main smoke value based on the dynamic zero point and the auxiliary smoke value based on the initial compensated zero point according to a third strategy; and activating or deactivating the smoke alarm based on the main smoke value and the auxiliary smoke value.

[0004] Existing technologies compensate for the temperature of smoke concentration collected by sensors or adjust the sensitivity of smoke sensors to avoid the influence of environmental factors. However, they rely on ambient temperature and humidity values ​​and ignore the instantaneous changes in temperature, humidity, and cooking fumes during cooking. For example, when steaming food, water vapor rises rapidly, and humidity increases significantly, leading to excessive changes in environmental factors during cooking. This causes the smoke concentration values ​​measured by the smoke sensor to be too high or too low, resulting in inaccurate data and making subsequent data analysis impossible. Summary of the Invention

[0005] The purpose of this invention is to solve the problem that existing technologies, which rely on static environmental characteristics for compensation, cannot adapt to various cooking kitchen environments, resulting in large deviations in smoke concentration measurements and affecting subsequent data analysis.

[0006] To achieve the above objectives, on the one hand, the present invention provides a smart integrated kitchen data acquisition method based on the Internet of Things, comprising:

[0007] The system acquires real-time smoke concentration data in the kitchen, corrects the real-time smoke concentration data to obtain corrected smoke concentration data, and uploads the corrected smoke concentration data to the cloud platform, thereby completing the data collection for the smart integrated kitchen.

[0008] The correction process includes: obtaining a pre-processed historical data sequence of the kitchen, the historical data sequence including historical smoke concentration time series and historical environmental feature time series, and constructing target sequences of historical smoke concentration time series and historical environmental feature time series respectively;

[0009] The historical smoke concentration time series, historical environmental feature time series, and corresponding target sequence are divided into windows according to a preset window length. Then, the static and dynamic correlations between environmental features and smoke concentration are calculated. The window length is calculated from the historical smoke concentration time series.

[0010] The compensation value of the real-time smoke concentration data is calculated based on static correlation and dynamic correlation. The corrected smoke concentration data is obtained by combining the compensation value and the real-time smoke concentration data.

[0011] By adopting the above technical solution to correct real-time smoke concentration data and combining it with historical kitchen data and correlation analysis of environmental characteristics, sensor errors and environmental interference can be effectively reduced, thereby providing more accurate smoke concentration data. This is of great significance for kitchen safety monitoring (such as fire early warning) and environmental management. This method utilizes IoT technology and data analysis algorithms to achieve dynamic monitoring and intelligent correction of the kitchen environment. Compared with traditional single-sensor monitoring, this method can better adapt to complex changes in the kitchen environment and improve the system's intelligence and adaptability. The corrected smoke concentration data is uploaded to the cloud platform for centralized management and analysis. This data management model not only improves data availability but also provides a foundation for subsequent big data analysis and intelligent decision-making.

[0012] Preferably, the environmental characteristics include temperature, humidity, and oil fume concentration in the kitchen.

[0013] By adopting the above technical solutions, environmental characteristics such as temperature, humidity, and oil fume concentration are included in the monitoring scope, which can more comprehensively reflect the actual environmental conditions of the kitchen, thereby providing a more accurate basis for the correction of smoke concentration. This multi-dimensional correction method can effectively reduce the error caused by single-factor correction and improve the accuracy of the corrected smoke concentration data.

[0014] Preferably, the construction process of the target sequence of historical smoke concentration time series is as follows:

[0015] The historical smoke concentration time series is processed by differential processing to obtain the differential value time series. The ratio of the differential value at each time point to the corresponding smoke concentration in the historical smoke concentration time series is calculated. All ratios are sorted in chronological order to obtain the target sequence of historical smoke concentration time series.

[0016] By adopting the above technical solution, differential processing can effectively capture the dynamic changes in smoke concentration. The differential value reflects the amount of change in smoke concentration between adjacent time points. By calculating the ratio of the differential value to the historical smoke concentration, the relative magnitude of this change can be further quantified, better reflecting the short-term fluctuations and trend changes in smoke concentration.

[0017] Preferably, the process for constructing the target sequence of the historical environmental feature time series is as follows:

[0018] The time series of historical environmental features is processed by differential processing to obtain the time series of differential values. The ratio of the differential value at each time step to the corresponding environmental feature in the time series of historical environmental features is calculated. All ratios are sorted in chronological order to obtain the target sequence of historical environmental feature time series.

[0019] By adopting the above technical solution, differential processing can capture the dynamic changes of environmental characteristics and reflect the changing trends between adjacent time points. This helps to identify short-term fluctuations in environmental characteristics, thereby more accurately reflecting the dynamic changes of the kitchen environment. Moreover, by calculating the ratio of the differential value to the historical environmental characteristics, the data is normalized. This normalization process can eliminate the influence of the absolute value on the data changing trend, making the data from different time periods comparable and facilitating subsequent analysis and modeling.

[0020] Preferably, the calculation process for the window length is as follows:

[0021] For historical smoke concentration time series, empirical mode decomposition is used to obtain multiple intrinsic mode function components. The Hilbert transform is used on each intrinsic mode function component to obtain the frequency of each intrinsic mode function component at each time step. The mean of the frequencies of a single intrinsic mode function component at all times is taken as the average frequency of that intrinsic mode function component. The reciprocal of the average frequency is taken as the window length, thus obtaining multiple window lengths.

[0022] By adopting the above technical solution, the reciprocal of the average frequency of each intrinsic mode function component is used as the window length. The window length can be dynamically adjusted according to the actual frequency characteristics of the data. This method avoids the errors that may be caused by a fixed window length, makes the window division more reasonable, and can better adapt to changes in the data. This allows the constructed model to learn and adapt to smoke concentration change patterns in different environments more effectively. The model can more accurately capture short-term fluctuations and long-term trends in the data, thereby improving the prediction accuracy of the model.

[0023] Preferably, the calculation process for the static correlation is as follows:

[0024] Based on any window length, the historical smoke concentration time series and the historical environmental feature time series are slid together to obtain multiple pairs of local sequences. Each pair of local sequences includes the historical smoke concentration time series and the historical environmental feature time series contained in each window sliding window.

[0025] Calculate the Pearson correlation coefficient for each pair of local sequences, and take the average of the Pearson correlation coefficients of all local sequence pairs as the static correlation value between environmental characteristics and smoke concentration under a single window length.

[0026] The average of the static correlation values ​​across all window lengths is taken as the static correlation.

[0027] By adopting the above technical solution and using a sliding window method based on different window lengths, it is possible to capture the local correlation between historical smoke concentration time series and historical environmental characteristic time series at different time scales. The static correlation value under each window length is calculated, and the average value is taken as the final static correlation. This method can comprehensively consider the correlation at different time scales. By calculating the Pearson correlation coefficient, the linear correlation between the two sequences can be quantified, while filtering out the influence of random noise, thus reflecting the intrinsic relationship between the data more accurately.

[0028] Preferably, the calculation process for the dynamic correlation is as follows:

[0029] Based on any window length, the target sequences corresponding to the historical smoke concentration time series and the historical environmental feature time series are slid together to obtain multiple pairs of local sequences. Each pair of local sequences includes the target sequence corresponding to the historical smoke concentration time series and the target sequence corresponding to the historical environmental feature time series contained in each window sliding window.

[0030] Calculate the Pearson correlation coefficient for each pair of local sequences, and take the average of the Pearson correlation coefficients of all local sequence pairs as the dynamic correlation value between environmental characteristics and smoke concentration under a single window length.

[0031] The average of the dynamic correlation values ​​across all window lengths is taken as the dynamic correlation.

[0032] Preferably, the calculation process for the compensation value is as follows:

[0033] Construct fitting curves between environmental features and the difference in smoke concentration, and between the target value of environmental features and the difference in smoke concentration. The target value of the environmental feature is the ratio of the difference in environmental features to the environmental feature.

[0034] Based on the fitted curve, the first smoke concentration difference value corresponding to the environmental characteristics at the current moment and the second smoke concentration difference value corresponding to the target value of the environmental characteristics at the current moment are calculated.

[0035] The compensation value is obtained by summing the product of the first smoke concentration difference and the static correlation, and the product of the second smoke concentration difference and the dynamic correlation.

[0036] By adopting the above technical solution, the first smoke concentration difference value and the second smoke concentration difference value obtained by fitting the curve can more accurately reflect the influence of environmental characteristics on smoke concentration at the current moment. Multiplying the static correlation and dynamic correlation with the corresponding difference values ​​and summing them can more comprehensively reflect the overall influence of environmental characteristics on smoke concentration, thereby improving the reliability of the compensation value.

[0037] Preferably, the specific calculation process for obtaining the corrected smoke concentration data by combining the compensation value and real-time smoke concentration data is as follows:

[0038] Calculate the average of the negative values ​​of the compensation values ​​for smoke concentration for all environmental features, and sum this average with the real-time smoke concentration data to obtain the corrected smoke concentration data.

[0039] By adopting the above technical solution, the influence of multiple environmental characteristics on smoke concentration can be comprehensively considered. This method can more comprehensively reflect the combined influence of environmental characteristics on smoke concentration, thereby improving the accuracy of the corrected smoke concentration data.

[0040] On the other hand, the present invention provides an IoT-based intelligent integrated kitchen data acquisition system, comprising: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the IoT-based intelligent integrated kitchen data acquisition method described in any of the above preferred technical solutions is implemented.

[0041] Compared with existing technologies, the beneficial effects of the IoT-based intelligent integrated kitchen data acquisition method and system of this invention are as follows:

[0042] 1. By correcting real-time smoke concentration data and combining it with historical kitchen data and environmental characteristics, correlation analysis can effectively reduce sensor errors and environmental interference, thereby providing more accurate smoke concentration data, which is of great significance for kitchen safety monitoring (such as fire early warning) and environmental management.

[0043] 2. This method utilizes IoT technology and data analysis algorithms to achieve dynamic monitoring and intelligent correction of the kitchen environment. Compared with traditional single-sensor monitoring, this method can better adapt to complex changes in the kitchen environment and improve the system's intelligence and adaptability.

[0044] 3. The corrected smoke concentration data is uploaded to the cloud platform for centralized management and analysis. This data management model not only improves the availability of data, but also provides a foundation for subsequent big data analysis and intelligent decision-making. Attached Figure Description

[0045] Figure 1 This is a flowchart of steps S1-S2 in the IoT-based smart integrated kitchen data acquisition method of this invention.

[0046] Figure 2 This is a flowchart of steps S10-S13 in the IoT-based smart integrated kitchen data acquisition method of this invention. Detailed Implementation

[0047] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0048] like Figure 1 As shown, the IoT-based smart integrated kitchen data acquisition method of the present invention includes:

[0049] S1: Acquire real-time smoke concentration data in the kitchen, correct the real-time smoke concentration data, and obtain corrected smoke concentration data.

[0050] like Figure 2 As shown, the process of correcting real-time smoke concentration data includes steps S10 to S13:

[0051] S10. Obtain the preprocessed historical data sequence of the kitchen. The historical data sequence includes the historical smoke concentration time series and the historical environmental feature time series. Construct the target sequences of the historical smoke concentration time series and the historical environmental feature time series respectively.

[0052] Because the kitchen environment is highly dynamic, factors such as temperature, humidity, airflow, and cooking fumes change differently at different cooking stages and with different food preparation methods. These changes in factors have varying impacts on the smoke sensor. Therefore, in one embodiment, environmental characteristics include temperature, humidity, and cooking fume concentration in the kitchen. A photoelectric smoke sensor is used to collect smoke concentration, a temperature sensor to collect temperature, a humidity sensor to collect humidity, and a cooking fume sensor to collect cooking fume concentration. Wavelet transform is used to denoise the collected data, yielding historical smoke concentration time series, historical temperature time series, historical humidity time series, and historical cooking fume concentration time series.

[0053] The construction process of the target sequence of historical smoke concentration time series is as follows: First-order differencing is performed on the historical smoke concentration time series to obtain a first-order differencing value time series. The ratio of the differencing value at each moment to the corresponding smoke concentration in the historical smoke concentration time series is calculated to obtain the target sequence of the historical smoke concentration time series. During first-order differencing, since the first moment in the historical smoke concentration time series has no preceding moment, differencing cannot be performed. Therefore, in this embodiment, the first-order differencing value corresponding to the first moment in the historical smoke concentration time series is defined as 0.

[0054] The construction process of the target sequence of historical temperature time series is as follows: First-order differencing is performed on the historical temperature time series to obtain first-order differencing value time series. The ratio of the differencing value at each moment to the corresponding temperature in the historical temperature time series is calculated to obtain the target sequence of the historical temperature time series. During first-order differencing, since the first moment in the historical temperature time series has no preceding moment, differencing cannot be performed. Therefore, in this embodiment, the first-order differencing value corresponding to the first moment in the historical temperature time series is defined as 0.

[0055] The construction process of the target sequences corresponding to historical humidity time series and historical oil fume concentration time series is the same as that of historical temperature time series, and will not be repeated here.

[0056] First-order differencing reveals the changing trends and dynamic characteristics of data by calculating the differences between adjacent data points in a historical data sequence. The first-order differencing value reflects the instantaneous change, and the ratio of the first-order differencing value to the historical data sequence represents the relative change, which is used for dynamic correlation analysis.

[0057] S11. Divide the historical smoke concentration time series, historical environmental feature time series and corresponding target sequence into windows according to the preset window length, and then calculate the static correlation and dynamic correlation between environmental features and smoke concentration. The window length is calculated from the historical smoke concentration time series.

[0058] The calculation process for the window length is as follows: For the historical smoke concentration time series, EMD (Empirical Mode Decomposition) is used to obtain N IMF (Intrinsic Mode Function) components. The Hippotransform is used on each IMF component to obtain the frequency of each IMF component at each time step. The mean of the frequencies of all times in a single IMF component is taken as the average frequency of that IMF component. The reciprocal of the average frequency is taken as the window length, thus obtaining N window lengths.

[0059] Using the reciprocal of the average frequency of each intrinsic mode function component as the window length allows for dynamic adjustment of the window length based on the actual frequency characteristics of the data. This method avoids the errors that may arise from a fixed window length, making the window division more reasonable and better adaptable to data changes. This enables the constructed model to learn and adapt more effectively to smoke concentration change patterns under different environments, and the model can more accurately capture short-term fluctuations and long-term trends in the data, thereby improving the model's prediction accuracy.

[0060] The calculation process for the static correlation between environmental characteristics and smoke concentration is as follows:

[0061] Since the calculation process for the static correlation between temperature, humidity, and oil fume concentration and smoke concentration is the same, to avoid redundancy, the following description uses the calculation process for the static correlation between temperature characteristics and smoke concentration as an example:

[0062] Based on window length The historical smoke concentration time series and historical temperature time series are divided into sliding windows to obtain multiple pairs of local sequences. Each pair of local sequences includes the historical smoke concentration time series and historical temperature time series contained in each sliding window.

[0063] The Pearson correlation coefficient for each pair of local sequences is calculated, and the average of the Pearson correlation coefficients for all local sequence pairs is taken as the static correlation value between temperature characteristics and smoke concentration under a single window length. Specifically, the formula for calculating the static correlation value is as follows:

[0064]

[0065] In the formula, This indicates that the window length is The static correlation values ​​between temperature and smoke concentration are divided by a sliding window. Indicates based on window length The number of windows obtained by dividing the sliding window. The first Pearson correlation coefficient for local sequences.

[0066] Calculate the static correlation between temperature and smoke concentration for other window lengths. The specific calculation process is the same as described above based on window length. The calculation process for partitioning is the same, so it will not be repeated here.

[0067] The static correlation between temperature and smoke concentration is obtained by averaging all static correlation values.

[0068] The calculation process for the dynamic correlation between environmental characteristics and smoke concentration is as follows:

[0069] Since the calculation process for the dynamic correlation between temperature, humidity, and oil fume concentration and smoke concentration is the same, to avoid redundancy, the following description uses the calculation process for the dynamic correlation between temperature characteristics and smoke concentration as an example:

[0070] Based on window length The target sequences corresponding to historical smoke concentration time series and historical temperature time series are divided into multiple pairs of local sequences by sliding window. Each pair of local sequences includes the target sequence corresponding to historical smoke concentration time series and the target sequence corresponding to historical temperature time series contained in each sliding window.

[0071] The Pearson correlation coefficient for each pair of local sequences is calculated, and the average of the Pearson correlation coefficients for all local sequence pairs is taken as the dynamic correlation value between temperature characteristics and smoke concentration under a single window length. Specifically, the formula for calculating the dynamic correlation value is as follows:

[0072]

[0073] In the formula, This indicates that the window length is The dynamic correlation between temperature and smoke concentration is divided by a sliding window. Indicates based on window length The number of windows obtained by dividing the sliding window. Indicates the first Pearson correlation coefficient for local sequences.

[0074] Calculate the dynamic correlation between temperature and smoke concentration for other window lengths. The specific calculation process is the same as described above based on window length. The calculation process for partitioning is the same, so it will not be repeated here.

[0075] The dynamic correlation between temperature and smoke concentration is obtained by averaging all dynamic correlation values.

[0076] Because cooking activities in the kitchen are complex, diverse, and rapidly changing, with varying durations for different activities, this invention, compared to traditional methods that analyze the correlation between overall features, uses a sliding window to consider the correlation between environmental features and smoke concentration during localized cooking processes. Different window lengths can be used to adapt to different cooking processes, enabling the capture of both short-term and long-term trends in cooking activities.

[0077] S12. Calculate the compensation value of real-time smoke concentration data based on static correlation and dynamic correlation.

[0078] Construct a fitting curve between environmental features and the difference in smoke concentration, with the x-axis representing the value of the environmental feature and the y-axis representing the difference in smoke concentration. Construct a fitting curve between the target value of the environmental feature and the difference in smoke concentration, with the target value of the environmental feature being the ratio of the difference in the environmental feature to the environmental feature value, with the x-axis representing the target value of the environmental feature and the y-axis representing the difference in smoke concentration.

[0079] Taking temperature characteristics as an example, a fitting curve for the temperature and smoke concentration difference is constructed based on the first-order difference time series of historical temperature and historical smoke concentration. The horizontal axis of the fitting curve is the temperature value, and the vertical axis is the smoke concentration difference value.

[0080] A fitting curve is constructed based on the target sequence corresponding to historical temperature time series and the first-order difference value time series of historical smoke concentration time series, corresponding to the temperature target value and the smoke concentration difference value. The horizontal axis of the fitting curve represents the temperature target value, and the vertical axis represents the smoke concentration difference value.

[0081] Based on the corresponding fitted curve, the first smoke concentration difference value corresponding to the current temperature value and the second smoke concentration difference value corresponding to the current temperature target value are calculated.

[0082] The product of the first smoke concentration difference and the static correlation, and the product of the second smoke concentration difference and the dynamic correlation are summed to obtain the temperature-induced smoke concentration compensation value. Specifically, the formula for calculating the temperature-induced smoke concentration compensation value is as follows:

[0083]

[0084] In the formula, This represents the compensation value for smoke concentration due to temperature. This represents the static correlation between temperature and smoke concentration. This represents the difference in smoke concentration corresponding to the current temperature value. This indicates the dynamic correlation between temperature and smoke concentration. This represents the difference in smoke concentration corresponding to the target temperature value at the current moment.

[0085] The impact of temperature changes during cooking should be reflected in the changes in smoke concentration. The smoke concentration difference value reflects the amount of change. Multiplying the amount of change corresponding to the temperature value by the static correlation, and multiplying the amount of change corresponding to the temperature target value by the dynamic correlation, can represent the actual contribution of temperature to the amount of change in smoke concentration.

[0086] Since smoke sensors typically measure smoke concentration based on the scattering of light beams from microparticles or the ionization of particles in the air, static temperature causes changes in air density, and instantaneous temperature changes cause instantaneous changes in the thermal motion of ions. Compared to traditional methods that compensate for smoke concentration based solely on static temperature, this invention achieves dual-modal smoke concentration compensation by considering both instantaneous temperature changes and static temperature. This allows for consideration of both the influence of static temperature on smoke concentration during cooking activities, thus compensating for both the effects of static temperature and the effects of temperature during cooking, making it adaptable to varying kitchen environments.

[0087] The calculation process for the compensation value of humidity and oil fume concentration to smoke concentration is the same as the calculation process for the compensation value of temperature to smoke concentration, and will not be repeated here.

[0088] S13. Combine the compensation value and real-time smoke concentration data to obtain the corrected smoke concentration data.

[0089] The average of the negative values ​​of the compensation values ​​for smoke concentration for all environmental features is calculated, and this average is summed with the real-time smoke concentration data to obtain the corrected smoke concentration data. Specifically, the formula for calculating the corrected smoke concentration data is as follows:

[0090]

[0091] In the formula, This represents the corrected smoke concentration value at the current moment. This indicates the real-time smoke concentration at the current moment collected by the smoke sensor. Indicates the first The compensation value of each environmental feature for smoke concentration.

[0092] Since the correlation range is (-1, 1), the compensation value should be subtracted for positively correlated features, and added to negatively correlated features. Taking temperature as an example, temperature is positively correlated with smoke concentration. When the temperature rises, the measured smoke concentration will be greater than the actual smoke concentration. Therefore, the compensation value of temperature for smoke concentration should be subtracted during correction to obtain the actual smoke concentration. Thus, the negative of the compensation value is used in the above formula for calculation.

[0093] S2: Upload the corrected smoke concentration data to the cloud platform to complete the data collection for the smart integrated kitchen.

[0094] After obtaining the corrected smoke concentration data, it can be uploaded to the cloud platform. The uploading process is based on existing technology and will not be described in detail here.

[0095] This completes the data collection for the smart integrated kitchen.

[0096] In addition, after uploading the corrected smoke concentration data to the cloud platform, it can be determined whether the corrected smoke concentration data exceeds the preset smoke concentration alarm threshold. If it exceeds the alarm threshold, an alarm is issued. The alarm threshold can be preset according to a conventional value range. For example, the alarm threshold can be set to 200-300 ppm, where ppm represents a concentration unit used to describe the proportion of a certain component in a mixture.

[0097] The IoT-based intelligent integrated kitchen data acquisition system includes a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement the IoT-based intelligent integrated kitchen data acquisition method described in the above embodiments.

[0098] The above description is only a preferred embodiment of the present invention. It should be noted that for ordinary counters in the art, several improvements and substitutions can be made without departing from the counting principle of the present invention, and these improvements and substitutions should also be considered within the scope of protection of the present invention.

Claims

1. A smart integrated kitchen data collection method based on the Internet of Things, characterized in that, The method comprises the following steps: Real-time smoke concentration data in the kitchen is acquired, and the real-time smoke concentration data is corrected to obtain corrected smoke concentration data; The corrected smoke concentration data is uploaded to a cloud platform, thereby completing intelligent integrated kitchen data acquisition; The correction process comprises the following steps: historical data sequences of the kitchen after preprocessing are acquired, the historical data sequences comprise historical smoke concentration time series and historical environmental feature time series, and target sequences of the historical smoke concentration time series and the historical environmental feature time series are respectively constructed; The historical smoke concentration time series, the historical environmental feature time series and the corresponding target sequences are windowed according to a preset window length, and then static correlation and dynamic correlation between the environmental feature and the smoke concentration are calculated, wherein the window length is calculated from the historical smoke concentration time series; A compensation value of the real-time smoke concentration data is calculated based on the static correlation and the dynamic correlation, and the corrected smoke concentration data is obtained by combining the compensation value and the real-time smoke concentration data; The construction process of the target sequences of the historical smoke concentration time series and the historical environmental feature time series is as follows: a difference value time series is obtained by differentiating the historical smoke concentration time series, the ratio of the difference value at each time to the corresponding smoke concentration in the historical smoke concentration time series is calculated, and the target sequence of the historical smoke concentration time series is obtained by sorting all the ratios in time sequence; A difference value time series is obtained by differentiating the historical environmental feature time series, the ratio of the difference value at each time to the corresponding environmental feature in the historical environmental feature time series is calculated, and the target sequence of the historical environmental feature time series is obtained by sorting all the ratios in time sequence; The static correlation is obtained from the historical smoke concentration time series and the historical environmental feature time series, the static correlation value between the environmental feature and the smoke concentration under a single window length is calculated, and the average value of the static correlation values under all window lengths is taken as the static correlation; The dynamic correlation is obtained from the target sequences corresponding to the historical smoke concentration time series and the historical environmental feature time series, the dynamic correlation value between the environmental feature and the smoke concentration under a single window length is calculated, and the average value of the dynamic correlation values under all window lengths is taken as the dynamic correlation. 2.The IoT-based smart integrated kitchen data collection method according to claim 1, wherein, The environmental feature comprises temperature, humidity and oil smoke concentration in the kitchen. 3.The IoT-based smart integrated kitchen data collection method of claim 1, wherein, The calculation process of the window length is as follows: For the historical smoke concentration time series, a plurality of intrinsic mode function components are obtained by using empirical mode decomposition, the frequency at each time in each intrinsic mode function component is obtained by using Hilbert transform, the average frequency of all frequencies at all times in a single intrinsic mode function component is taken as the average frequency of the intrinsic mode function component, the reciprocal of the average frequency is taken as the window length, and thus a plurality of window lengths are obtained. 4.The IoT-based smart integrated kitchen data collection method of claim 3, wherein, The calculation process of the static correlation is as follows: The historical smoke concentration time series and the historical environmental feature time series are respectively slid based on any window length, a plurality of pairs of local sequences are obtained, each pair of local sequences comprises the historical smoke concentration time series and the historical environmental feature time series contained in the window in each window sliding; The Pearson correlation coefficients of each pair of local sequences are calculated, and the average value of the Pearson correlation coefficients of all pairs of local sequences is taken as the static correlation value between the environmental feature and the smoke concentration under a single window length. The average of the static correlation values under all window lengths is taken as the static correlation. 5.The IoT-based smart integrated kitchen data collection method of claim 4, wherein, The calculation process of the dynamic correlation is as follows: The target sequences corresponding to the historical smoke concentration time sequence and the historical environmental feature time sequence are respectively slid based on any window length, and a plurality of pairs of local sequences are obtained, each pair of local sequences including a target sequence corresponding to the historical smoke concentration time sequence and a target sequence corresponding to the historical environmental feature time sequence contained in the window in each window sliding; The Pearson correlation coefficient of each pair of local sequences is calculated, and the average of the Pearson correlation coefficients of all pairs of local sequences is taken as the dynamic correlation value of the environmental feature and the smoke concentration under a single window length. The average of the dynamic correlation values under all window lengths is taken as the dynamic correlation. 6.The IoT-based smart integrated kitchen data collection method of claim 1, wherein, The calculation process of the compensation value is as follows: A fitting curve of the environmental feature and the smoke concentration difference value and a fitting curve of the target value of the environmental feature and the smoke concentration difference value are constructed, the target value of the environmental feature being the ratio of the difference value of the environmental feature to the environmental feature; The first smoke concentration difference value corresponding to the current moment of the environmental feature and the second smoke concentration difference value corresponding to the target value of the current moment of the environmental feature are calculated based on the fitting curve; The product of the first smoke concentration difference value and the static correlation and the product of the second smoke concentration difference value and the dynamic correlation are summed to obtain the compensation value. 7.The IoT-based smart integrated kitchen data collection method according to claim 1, wherein, The specific calculation process of the modified smoke concentration data obtained by combining the compensation value and the real-time smoke concentration data is as follows: The average of the reciprocals of all environmental feature compensation values of the smoke concentration is calculated, and the average is summed with the real-time smoke concentration data to obtain the modified smoke concentration data.

8. The intelligent integrated kitchen data acquisition system based on the Internet of Things, characterized in that, It comprises: A processor and a memory, the memory stores computer program instructions, when the computer program instructions are executed by the processor, the method for collecting data of intelligent integrated kitchen based on Internet of Things according to any one of claims 1-7 is realized.

Images