A data acquisition method and system based on IoT devices

Abstract

The present invention relates to the field of data processing technology, and more specifically to a data collection method and system based on IoT devices. The method comprises: installing smoke concentration sensors and temperature sensors in the cargo area of a warehouse; determining a smoke concentration risk index and a temperature risk index based on the deviation degree and rate of change of the current smoke concentration and temperature, respectively; and determining a comprehensive risk index based on the mutual information between the smoke concentration and temperature; determining a comprehensive value index based on the uncertainty of historical smoke concentration and temperature series and their similarity to corresponding standard series; and utilizing the comprehensive risk index and comprehensive value index to adjust the sensor collection frequency, thereby implementing an adaptive data collection strategy. This method can promptly respond to environmental changes and identify potential risks while avoiding the generation of excessive redundant data.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and more particularly to a data acquisition method and system based on an Internet of Things device. Background Art

[0002] With the widespread adoption of the Internet of Things (IoT), data acquisition systems based on various sensors have found applications in warehouse security, industrial monitoring, environmental monitoring, and other scenarios. In these scenarios, temperature and smoke concentration sensors, as core sensing devices for warehouse fire monitoring, are widely used due to their simple mechanisms and low deployment costs.

[0003] In the prior art, temperature sensors and smoke concentration sensors usually use a fixed acquisition frequency for data collection, that is, they periodically sample environmental data within a set time interval and trigger an alarm when the detection value exceeds a threshold.

[0004] However, in practice, warehouse environments are highly dynamic, and fire outbreaks are characterized by suddenness and nonlinearity. Fixed data collection frequencies cannot adapt to rapidly changing environmental conditions. This is especially true during periods of rapidly evolving fire risk. If the fixed data collection frequency is set too low, the rate of change in temperature and smoke concentration will far exceed the timescale within which the fixed data collection frequency can detect. This can easily lead to problems such as delayed risk identification and untimely alarm responses. Setting a high, continuous data collection frequency will also generate large amounts of redundant data during periods of stable conditions or low risk, leading to data storage pressure, wasted computing resources, and increased communication load.

[0005] Therefore, there is an urgent need for a method that can adjust the data collection strategy in real time according to the dynamic changes of environmental conditions, which can not only respond to environmental changes in a timely manner and effectively identify potential risks, but also avoid the generation of excessive redundant data. Summary of the Invention

[0006] In order to solve the problems of data redundancy and risk lag caused by the fixed collection frequency in the prior art, the present invention proposes a data collection method and system based on Internet of Things devices.

[0007] In a first aspect, the present invention provides a data collection method based on an Internet of Things device, comprising:

[0008] Install smoke concentration sensors and temperature sensors in the cargo area of the warehouse;

[0009] Taking any moment as the current moment, determine the smoke concentration risk index and temperature risk index at the current moment based on the respective deviation degrees of the smoke concentration and temperature at the current moment, as well as the respective change rates of the smoke concentration and temperature at the current moment;

[0010] Determining the mutual information of the smoke concentration and temperature at the current moment based on the historical smoke concentration sequence and the historical temperature sequence at the current moment, and determining the comprehensive risk index of the smoke concentration and temperature at the current moment by combining the mutual information with the smoke concentration risk index and the temperature risk index;

[0011] Determine the comprehensive value index of the current smoke concentration and temperature based on the uncertainty of the historical smoke concentration series and the historical temperature series and their similarity with the corresponding standard series; the uncertainty is determined based on the information entropy value;

[0012] The comprehensive risk index and the comprehensive value index are used to dynamically adjust the acquisition frequencies of the smoke concentration sensor and the temperature sensor to achieve an adaptive data acquisition strategy.

[0013] This technical solution, by deploying smoke concentration and temperature sensors in the cargo area, enables real-time monitoring of environmental changes, providing accurate data support for data collection and subsequent risk identification. By analyzing the deviation and rate of change of current smoke concentration and temperature from historical data, it can identify abnormal data fluctuations and trends, and then calculate a corresponding risk index, providing real-time feedback on potential environmental risks. By comprehensively considering the correlation between smoke concentration and temperature, a comprehensive risk index is determined, thereby improving the accuracy of risk assessment. Furthermore, by incorporating information entropy to calculate the uncertainty of historical data, the method effectively reflects the stability or volatility of the data. Combining the uncertainty of historical data and its similarity to standard data, it comprehensively assesses the contribution of real-time data to risk assessment. This comprehensive value index helps the system make more precise adjustments and optimizes the data collection strategy. Finally, by adjusting the collection frequency in real time, the data collection strategy can respond to environmental changes in real time, avoiding the lag and redundant data associated with fixed-frequency collection, ensuring efficient system operation and rational resource utilization.

[0014] Furthermore, the dynamic adjustment of the acquisition frequency of the smoke concentration sensor and the temperature sensor is based on the following formula:

[0015]

[0016] In the formula, The adjusted acquisition frequency for the smoke density sensor or temperature sensor. is the initial acquisition frequency of the smoke density sensor or temperature sensor, for The comprehensive risk index of temperature and smoke concentration at the time, is a symbolic function, for The comprehensive value index of temperature and smoke density at the moment, for The theoretical median value of and It corresponds to the smoke density sensor or the temperature sensor at the same time.

[0017] This technical solution closely aligns the data collection frequency of smoke and temperature sensors with the current environmental risk and data value. This ensures that data collection frequency is increased when risk is high to capture environmental changes promptly, while frequency is reduced when risk is low to avoid redundant data generation. By combining a comprehensive risk index with a value index and using a sign function to control the direction of frequency adjustment, the system can flexibly respond to environmental changes, improving responsiveness while optimizing resources.

[0018] Furthermore, dynamically adjusting the acquisition frequencies of the smoke concentration sensor and the temperature sensor also includes adding boundary limits to the adjusted acquisition frequencies of the smoke concentration sensor and the temperature sensor, including: obtaining the respective maximum acquisition frequency and minimum acquisition frequency based on the respective hardware parameters of the smoke concentration sensor and the temperature sensor; if the adjusted acquisition frequency of the smoke concentration sensor or the temperature sensor is less than the respective minimum acquisition frequency, setting the adjusted acquisition frequency to the respective minimum acquisition frequency; if the adjusted acquisition frequency of the smoke concentration sensor or the temperature sensor is greater than the respective minimum acquisition frequency, setting the adjusted acquisition frequency to the respective maximum acquisition frequency.

[0019] This technical solution sets hardware-level maximum and minimum boundary limits for the acquisition frequency of the smoke concentration sensor and temperature sensor, ensuring that the acquisition frequency of the smoke concentration sensor and the temperature sensor is always within the physical limitations of the device. This improves the rationality of dynamically adjusting the acquisition frequency and enables the system to ensure the stability and long-term reliability of the sensor when dynamically adjusting the frequency.

[0020] Furthermore, the comprehensive risk index of smoke concentration and temperature at the current moment is determined based on the following formula:

[0021]

[0022] In the formula, for The comprehensive risk index of smoke concentration and temperature at the moment, for The smoke concentration risk index at the moment, for Temperature risk index at the moment, for The weight of the smoke concentration risk index at the moment, for The weight of the temperature risk index at the moment, is the normalization function, is the mutual information between smoke concentration and temperature at the current moment, for The information entropy value of the historical smoke concentration sequence at time for The information entropy value of the historical temperature series at the time.

[0023] This technical solution comprehensively considers the risk factors of smoke concentration and temperature, and through weight adjustment, normalization functions, and mutual information calculation, accurately assesses the current environmental risk level. Combining the smoke concentration and temperature risk index with the stability and correlation of historical data, it comprehensively reflects the complexity of environmental changes and the multidimensional nature of risk. This not only improves the accuracy and sensitivity of the risk index, but also enhances the real-time response to environmental changes.

[0024] Furthermore, the comprehensive value index of smoke concentration and temperature at the current moment is determined based on the following formula;

[0025]

[0026] In the formula, for The comprehensive value index of smoke concentration and temperature at the moment, for The information entropy value of the historical smoke concentration sequence at time for The similarity between the historical smoke concentration sequence and the standard smoke concentration sequence at the moment for The information entropy of the historical temperature sequence at the time, for The similarity between the historical temperature series and the standard temperature series at the moment.

[0027] This technical solution comprehensively assesses the actual value of smoke concentration and temperature data based on the uncertainty of historical data and its similarity to standard data. The stability of historical data, combined with its similarity to standard data, dynamically reflects the credibility of real-time data and its contribution to risk assessment. This comprehensive value index effectively identifies data with a high contribution to risk assessment.

[0028] Furthermore, the smoke concentration risk index and temperature risk index at the current moment are determined based on the following formula:

[0029]

[0030] In the formula, for The risk index of smoke concentration or temperature at the moment, for The degree of deviation of smoke concentration or temperature at the moment, for The degree of deviation of smoke concentration or temperature at the moment, To take the absolute value sign, for The rate of change of smoke concentration or temperature at a given moment, is the median of the rate of change of the historical smoke concentration series or historical temperature series at time t; 、 、 、 and It corresponds to the smoke concentration or temperature at the same time.

[0031] This technical solution comprehensively assesses the potential risks of the environment by defining risk indices for smoke concentration and temperature. It combines the degree of deviation and rate of change between the current moment and the previous moment, and through nonlinear weighting of the degree of deviation and rate of change, makes the risk index more sensitive to sudden changes or abnormal fluctuations, helping to promptly identify potential risks in the environment, especially in scenarios with drastic changes.

[0032] Furthermore, the weight of the smoke concentration risk index and the weight of the temperature risk index are determined based on the following method: calculating the cumulative value of the coefficient of variation of the historical smoke concentration sequence and the coefficient of variation of the historical temperature sequence at the current moment; taking the ratio of the coefficient of variation of the historical smoke concentration sequence at the current moment to the cumulative value as the weight of the smoke concentration risk index at the current moment; taking the ratio of the coefficient of variation of the historical temperature sequence at the current moment to the cumulative value as the weight of the temperature risk index at the current moment.

[0033] Furthermore, the historical smoke concentration sequence and the historical temperature sequence are determined as follows: all historical smoke concentrations and all temperatures within a preset time period before the current moment constitute the historical smoke concentration sequence and the historical temperature sequence at the current moment, respectively.

[0034] Furthermore, the standard smoke concentration sequence and the standard temperature sequence are determined based on the following method: the smoke concentration and temperature of the cargo area under normal conditions within a preset time period are pre-collected to form the standard smoke concentration sequence and the standard temperature sequence, respectively; and the preset time period is the same length as the preset time period when obtaining the historical smoke concentration sequence and the historical temperature sequence.

[0035] In a second aspect, the present invention provides a data acquisition system based on Internet of Things devices, comprising: a temperature sensor and a smoke concentration sensor, respectively installed in the cargo area of the warehouse; a control unit, integrated with a computer program, which, when executed, responds to the data collected by the smoke concentration sensor and the temperature sensor, and executes the steps of the data acquisition method to implement an adaptive data acquisition strategy.

[0036] The present invention has the following effects:

[0037] This invention adaptively adjusts the data collection frequency based on real-time changes in environmental conditions, enabling efficient response and accurate perception of environmental changes. In emergencies like fires, the system automatically increases the collection frequency to ensure a timely response. When the risk is low or the environment is stable, the system automatically decreases the collection frequency, reducing data redundancy and resource waste. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a schematic flow chart of the method of the present invention;

[0039] Figure 2 1 is a schematic flow chart of the method of step S2 of the present invention. DETAILED DESCRIPTION

[0040] The technical solutions in the embodiments of the present invention will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present invention.

[0041] The present invention provides a data collection method based on IoT devices, such as Figure 1 As shown in , including:

[0042] S1: Obtain data from temperature sensors and smoke concentration sensors in each cargo area of the warehouse.

[0043] The warehouse is divided into multiple cargo areas, and a temperature sensor and a smoke concentration sensor are installed in each cargo area. The smoke concentration sensor usually collects the concentration of particulate matter or combustible gas in the air. Specifically, a photoelectric smoke concentration sensor is used to collect the smoke concentration in the air, and the unit is usually ppm. The temperature sensor usually collects the ambient temperature, and the unit is .

[0044] For any cargo area, according to the traditional collection method, the smoke concentration sensor and temperature sensor are set to a fixed collection frequency of 1Hz (collection once per second), and the smoke concentration and temperature of the cargo area are obtained in real time. At this moment, the cargo area is The corresponding smoke concentration at the time is , the temperature is .

[0045] S2: Based on the data from the temperature sensor and the smoke concentration sensor, the comprehensive risk index of the smoke concentration and temperature at the current moment is evaluated in real time.

[0046] In warehouse security monitoring, the real-time assessment of a comprehensive risk index is crucial for early identification and dynamic response to sudden risks like fires. Warehouses store large quantities of goods, and temperature and smoke fluctuations are key precursors to fire development. By analyzing data from temperature and smoke sensors in real time and constructing a comprehensive risk index, we can more accurately reflect the current level of danger in the environment.

[0047] For example, when the smoke concentration suddenly rises or the temperature fluctuates violently, especially when both change significantly at the same time and have a certain synergistic trend, it is very likely to indicate the initial formation or spread of a fire source, and timely early warning processing is needed to resolve the safety hazards in the warehouse.

[0048] This step quantifies risk by integrating the degree of data deviation, the rate of change of data, and the correlation between modalities to obtain a comprehensive risk index for the current smoke concentration and temperature. This can achieve more fine-grained, physically meaningful real-time perception, providing an accurate basis for subsequent acquisition frequency adjustment and early warning linkage.

[0049] Specifically, first obtain the temperature and smoke concentration within N seconds (the empirical value is 60) before the current moment, and obtain the historical temperature series and historical smoke concentration series at the current moment. Then, obtain the comprehensive risk index of the smoke concentration and temperature at the current moment, such as Figure 2 As shown in , including:

[0050] S21: Determine the current smoke concentration deviation and temperature deviation.

[0051] The deviation of the current smoke concentration is measured by dividing the absolute value of the difference between the current smoke concentration and the mean of the historical smoke concentration series by the standard deviation of the historical smoke concentration series.

[0052] The specific formula is:

[0053]

[0054] In this formula, for The degree of deviation of the smoke concentration at the moment (standardized deviation, normalized result using standard deviation), for The smoke concentration at the time, for The mean of the historical smoke concentration series at time , for The standard deviation of the historical smoke concentration series at time t.

[0055] The degree of deviation of the temperature at the current moment is measured by dividing the absolute value of the difference between the current temperature and the mean of the historical temperature series (at the current moment) by the standard deviation of the historical temperature series.

[0056] The specific formula is:

[0057]

[0058] In the formula, for The degree of temperature deviation at the moment, for The temperature of the moment, for The mean of the historical temperature series at time , for The standard deviation of the historical temperature series at time t, where || is the absolute value symbol.

[0059] This method for calculating data deviation is essentially a Z-score (standard score), an effective way to measure the degree of difference between current data and historical data. Specifically, it normalizes the difference between the current data and the historical mean, taking into account data volatility (standard deviation), thereby eliminating the influence of data units and magnitude, making data from different times and devices comparable. In warehouse monitoring scenarios, a significant deviation in smoke concentration or temperature at a given moment often indicates a potential anomaly or emergency, such as a fire or fever causing the temperature to deviate from the historical average, or smoke accumulation causing the smoke concentration to deviate from the historical average.

[0060] S22: Evaluate the smoke concentration risk index and temperature risk index based on the respective change rates of the smoke concentration and temperature at the current moment.

[0061] The current smoke concentration change rate is determined by calculating the difference between the current and previous smoke concentrations, reflecting the speed of smoke concentration change. It is a crucial real-time feedback indicator in monitoring systems, especially in scenarios with potential safety risks such as fires or hazardous gas leaks. The rate of change helps detect anomalies promptly and prevent further incidents. A high smoke concentration change rate may indicate a sudden fire or equipment failure.

[0062] Specifically, the calculation formula is:

[0063]

[0064] In the formula, yes The temperature change rate at time yes The temperature of the moment, for The temperature of the moment, is an absolute value.

[0065] The current temperature rate of change is determined by calculating the temperature difference between the current moment and the previous moment (or within a historical period). It reflects the speed of temperature change. It is an important real-time feedback indicator in monitoring systems, especially in scenarios where sudden temperature changes may cause safety issues (such as equipment overheating and fire). The rate of change helps detect potential risks early.

[0066] Specifically, the calculation formula is:

[0067]

[0068] In the formula, yes The rate of change of smoke concentration at the moment, yes The smoke concentration at the time, for The smoke concentration at the time, is an absolute value.

[0069] Next, using the same method, all smoke concentration change rates corresponding to the historical smoke concentration sequence at the current moment and all temperature change rates corresponding to the historical temperature sequence at the current moment are obtained.

[0070] Subsequently, the smoke concentration risk index at the current moment is calculated based on the deviation degree of the smoke concentration and the smoke concentration change rate at the current moment, and the temperature risk index at the current moment is calculated based on the deviation degree of the temperature and the temperature change rate at the current moment.

[0071] The core logic of the calculation here is: the greater the deviation of a data from historical data and the faster the rate of change of the data, the more likely there is risk at the moment corresponding to the data, and the greater the risk index of the data, and vice versa.

[0072] In one embodiment, the smoke concentration risk index at the current moment is determined based on the following formula:

[0073]

[0074] In this formula, for The risk index of smoke concentration at the moment, for The degree of deviation of smoke concentration at the moment, for The degree of deviation of smoke concentration at the moment, To take the absolute value sign, for The rate of change of smoke concentration at the moment, for The median of the rate of change of the historical smoke concentration series at the time instant.

[0075] In this formula, It combines the current smoke concentration deviation and the variation of the current smoke concentration deviation. If the current smoke concentration differs more from the average level of historical smoke concentration, that is, the deviation is greater, and the variation of the deviation is also greater, then the smoke concentration at the current moment is more likely to be risky. Dividing by 2 takes into account that the two are equally important, which is equivalent to giving them the same weight. The main purpose of squaring is to amplify the impact of the smoke concentration deviation and its variation. In fire risk assessment, abnormal deviation and rapid change of smoke concentration are both critical signals. Squaring can make these abnormal conditions more prominent in the risk score.

[0076] In this formula, The "%" portion represents the current rate of change of smoke concentration relative to the historical average. Adding 1 to this portion emphasizes the difference in the current rate of change of smoke concentration relative to the historical average without significantly altering the original magnitude. This serves to determine whether the current rate of change is faster or slower than the historical average. If the current rate of change is significantly greater than the historical average, it indicates that smoke concentration is changing rapidly, potentially indicating an abnormal situation such as a fire. This portion incorporates the impact of this rapid change into the risk score, improving the timeliness and accuracy of risk assessments.

[0077] This design allows both calculated values to be larger when the smoke concentration deviates significantly and the rate of change is faster. Multiplying them together significantly increases the risk index, highlighting the high-risk situation. Conversely, if the deviation is small and the rate of change is slow, the multiplication results in a lower risk index, reflecting a low-risk situation. Multiplication allows the two risk factors to interact, making the smoke concentration risk index more consistent with the actual risk situation.

[0078] In one embodiment, the temperature risk index at the current moment is determined based on the following formula:

[0079]

[0080] In the formula, for Temperature risk index at the moment, for The degree of temperature deviation at the moment, for The degree of temperature deviation at the moment, To take the absolute value sign, for The temperature change rate at time for The median of the rate of change of the historical temperature series at the moment.

[0081] In this formula, The system first combines the current temperature deviation and the magnitude of the temperature deviation. The deviation measures the difference between the current temperature and the historical average. Larger deviations often indicate potential anomalies. If the magnitude of the deviation is also large, it indicates that the abnormal temperature fluctuation may be more severe, thereby increasing the risk level at the current moment. By calculating the sum of the deviation and the magnitude of the deviation and dividing it by 2, the importance of both is guaranteed to be equal. The purpose of the squaring operation is to amplify the impact of the temperature deviation and its magnitude on the risk assessment. In fire and overheating scenarios, extreme temperature deviations (especially drastic temperature changes) often indicate serious safety risks. The squaring operation makes these significant abnormal changes have a stronger amplifying effect on the risk index, further increasing the system's sensitivity to risk.

[0082] In this formula, The current temperature change rate is measured relative to the median of the historical temperature change rates. This ratio determines how fast the current temperature change rate compares to the historical temperature change rate. If the current temperature change rate is significantly greater than the historical temperature change rate, it indicates a rapid temperature change, potentially indicating a dangerous situation such as equipment overheating or fire. Adding 1 ensures calculation stability and incorporates the difference between the current temperature change rate and the historical level into the risk assessment, effectively increasing sensitivity to rapid temperature changes.

[0083] With this design, when both the degree of temperature deviation and the rate of change are large, the results of both calculations will be larger, and the final risk index will also increase significantly. This highlights the risk of conditions with more drastic temperature changes (such as overheating or fire). This design makes dangerous situations like fire easier to detect, thereby improving the system's response speed and accuracy. When the temperature changes more slowly and the degree of deviation is smaller, the risk index will be lower, which appropriately reflects low-risk conditions, helps avoid over-response, and ensures that the system does not frequently issue unnecessary alarms under normal circumstances.

[0084] S23: The mutual information of smoke concentration and temperature is introduced, and the comprehensive risk index is determined by combining the risk indices of smoke concentration and temperature.

[0085] In a fire scenario, while smoke concentration and temperature typically change as the fire develops, the relationship between them is not a simple linear one. By calculating mutual information, we can understand the statistical correlation between these two variables. A high mutual information between two variables indicates a strong synergistic effect in fire risk assessment. Therefore, a more comprehensive consideration of this interrelationship is necessary when calculating the comprehensive risk index, thereby more accurately reflecting the actual fire risk situation and improving the accuracy of risk assessment.

[0086] Specifically, they include:

[0087] S231: Obtain the mutual information between the temperature data and the smoke data at the current moment.

[0088] Normalization is performed on the historical smoke concentration series and the historical temperature series at the current moment. Specifically, the maximum and minimum value normalization method is used to normalize the values of each smoke concentration in the historical smoke concentration series and each temperature in the historical temperature series to Within the range.

[0089] The value ranges of historical smoke concentration series and historical temperature series are divided into The intervals of historical smoke concentration data and historical temperature data are combined into equal intervals. interval combinations;

[0090] In the historical smoke concentration series and historical temperature series, the smoke concentration and temperature at the same moment are regarded as a pair of data combinations. According to the number of occurrences of each pair of data combinations in each interval combination, the joint probability distribution and marginal probability distribution are statistically analyzed. Based on the joint probability distribution and marginal probability distribution, the mutual information between smoke concentration and temperature is calculated, which is then used as the mutual information between temperature and smoke at the current moment.

[0091] For example, the numerical intervals of the historical smoke concentration series and the historical temperature series after division are both:

[0092] [0,0.2), [0.2,0.4), [0.4,0.6), [0.6,0.8), [0.8,1.0]; are respectively recorded as interval 1 to interval 5, then the values of the historical smoke concentration series are divided into interval 1 to interval 5, and the values of the historical temperature series are also divided into interval 1 to interval 5.

[0093] In the historical smoke concentration series and historical temperature series, The smoke concentration and temperature corresponding to the time are and , It is mapped to the smoke concentration interval 3, recorded as C3, is mapped to temperature interval 4, recorded as T4, then and Constitute a pair of data combinations, the interval combination that this pair of data combinations falls into is .

[0094] Statistical joint probability distribution: Joint probability distribution describes the probability of different combinations of smoke concentration and temperature. If the length of the historical smoke concentration series and the historical temperature series is 60, that is, 60 moments, a total of 60 pairs of data combinations will be obtained (without deduplication). If 5 of the 60 pairs of data combinations fall into , then the statistical joint probability distribution is:

[0095]

[0096] In this formula, for The joint probability of this interval combination indicates that in the historical smoke concentration series and the historical temperature series, the probability of the data combination of smoke concentration in the C3 interval and temperature in the T4 interval occurring is 0.0833. In a similar way, the joint probability of all 25 interval combinations can be calculated.

[0097] Statistical marginal probability distribution: The marginal probability distribution represents the probability of a single variable (such as smoke concentration or temperature) under all possible values, and the marginal probability distribution is calculated from the joint probability distribution.

[0098] The marginal probability of smoke concentration in the C3 interval is , here 0 means the 0 temperature intervals, express The joint probability of this interval combination is, 0 ranges from 1 to 5, and the smoke concentration is obtained by accumulating the joint probability. The marginal probability of the interval.

[0099] The marginal probability of the temperature being in the T4 interval is , here Indicates the Smoke concentration range, express The joint probability of this interval combination is obtained by accumulating the joint probability. Marginal probabilities in the interval.

[0100] Through the joint probability distribution and marginal probability distribution, the mutual information between the smoke concentration and temperature at the current moment can be calculated:

[0101]

[0102] In this formula, is the mutual information between smoke concentration and temperature, reflecting the correlation between the two variables. and are all interval numbers. The smoke concentration interval and temperature interval The interval combination The joint probability of The smoke concentration is The marginal probability in the interval, The temperature is The marginal probability in the interval.

[0103] Since mutual information reflects the correlation between two variables, the mutual information between the smoke concentration and temperature at the current moment is , whose value is calculated based on the historical smoke concentration series and temperature series at the current moment .

[0104] Calculating mutual information is a dynamic way to reflect the coordinated change trend between smoke concentration and temperature at the current moment. Specifically, the stronger the coordinated change trend between smoke concentration and temperature over the past 60 seconds, the greater the mutual information value at the current moment, indicating that the smoke concentration and temperature are more likely to increase together. If the coordinated change trend between smoke concentration and temperature is weaker, the smaller the mutual information value, indicating that the changes in smoke concentration and temperature are not driven by the same factor or their mutual influence is relatively weak, and the smoke concentration and temperature may be relatively independent, which generally means that the system is in a normal and stable operating state.

[0105] S232: Determine the comprehensive risk index of smoke concentration and temperature at the current moment.

[0106] Under normal circumstances, smoke concentrations typically remain relatively low and stable. They may fluctuate slightly due to factors like dust and ventilation, but generally do not experience sustained increases or significant changes. Temperature fluctuations are similar. Under normal circumstances, they exhibit a certain regularity depending on the environment, such as relatively stable fluctuations with the alternation of day and night and the change of seasons, and the fluctuations are generally within a certain range.

[0107] When a fire breaks out, smoke concentration often rises rapidly, serving as a key indicator of the fire's early stages. Even small fires can produce large amounts of smoke, and this change is characterized by suddenness and rapid growth. Temperature also gradually rises after a fire breaks out, but the rate of increase is relatively slow and influenced by various factors, including environmental factors and fire type. Unlike smoke concentration, it can clearly and quickly reflect the fire's status in its early stages.

[0108] Therefore, when calculating the comprehensive risk index, a greater weight is assigned to smoke concentration and a smaller weight is assigned to temperature. This is because smoke concentration is a more sensitive and direct indicator of fire in the early stages of a fire. Its sudden increase is crucial for fire early warning, buying more time for timely fire detection and intervention. Temperature changes, on the other hand, may be less significant in the early stages of a fire and their indicative effect is delayed. Therefore, a relatively smaller weight is assigned to temperature. This allows for a more accurate assessment of fire risk through the comprehensive risk index, improving the timeliness and accuracy of early warnings.

[0109] In one embodiment, the weights are set based on the coefficient of variation of the historical smoke concentration series and the coefficient of variation of the historical temperature series at the current moment. Specifically:

[0110] Calculate the coefficient of variation of the historical smoke concentration series at the current moment , Equal to the standard deviation of the historical smoke concentration series divided by the mean of the historical smoke concentration series, and calculate the coefficient of variation of the historical temperature series at the current moment , It is equal to the standard deviation of the historical temperature series divided by the mean of the historical temperature series.

[0111] Under normal circumstances, the coefficient of variation of historical smoke concentration and temperature series is typically small, remaining within a relatively stable range. However, during a fire, smoke is a product of the combustion process. Due to the complexity and uncertainty of factors such as combustion intensity and the burning material, the dispersion of smoke concentration data increases significantly, and the coefficient of variation of smoke concentration data increases significantly. Temperature also rises rapidly, but the rise is relatively continuous, unlike the sudden and drastic fluctuations of smoke concentration. Therefore, the dispersion of temperature changes less than that of smoke concentration. In other words, the increase in the coefficient of variation of temperature is usually smaller than that of smoke concentration.

[0112] In summary, the comprehensive risk index of smoke concentration and temperature at the current moment is determined based on the following formula:

[0113]

[0114] In this formula, for The comprehensive risk index of smoke concentration and temperature at the moment, for The smoke concentration risk index at the moment, for Temperature risk index at the moment, for The weight of the smoke concentration risk index at the moment, for The weight of the temperature risk index at the moment, is the normalization function, since is a positive number greater than 0, and the range of the normalized function in the range from 0 to positive infinity is , by multiplying by 2 and then subtracting 1, we get The value of interval, to achieve Normalization of . is the mutual information between smoke concentration and temperature at the current moment, for The information entropy value of the historical smoke concentration sequence at time for The information entropy value of the historical temperature series at the time.

[0115] In this formula, , ,in, is the coefficient of variation of the historical smoke concentration series at the current moment, is the coefficient of variation of the historical temperature series at the previous moment. This is because if a fire occurs, the coefficient of variation of smoke concentration will change more significantly, so a larger weight is given to the risk index of smoke concentration. The change in the coefficient of variation of temperature is relatively less obvious, so a smaller weight is given to the risk index of temperature.

[0116] In this formula, Part of it is a common way to normalize mutual information so that its value is in the range of 0 to 1. for The information entropy value of the historical smoke concentration sequence at time for The information entropy value of the historical temperature series at the time.

[0117] In summary, this formula reflects the degree of correlation between smoke concentration and temperature through mutual information. When the changing trends of the two are coordinated and the higher the risk index of each is, the larger the comprehensive risk index is and the closer it is to 1, it means that a fire is more likely to occur at the current moment and there is a high fire risk, and vice versa.

[0118] S3: Based on the data from the temperature sensor and the smoke density sensor, the temperature and smoke density data under standard conditions are compared to evaluate the comprehensive value index of the smoke density and temperature at the current moment.

[0119] In fire risk assessment, relying solely on risk indices based on temperature and smoke concentration may not fully reflect the potential risk level, as these indicators can be affected by environmental fluctuations and exhibit temporary anomalies. Therefore, it is necessary to analyze data anomalies and uncertainties to quantify the data's value index. This value index can more accurately reflect potential fire risks and help more effectively assess and respond to potential crises.

[0120] Specifically, they include:

[0121] Get the information entropy of the historical smoke concentration sequence at the current moment and the information entropy of the historical temperature series , to measure the uncertainty of historical smoke concentration series and historical temperature series. The larger the information entropy value, the greater the uncertainty, and the higher the potential value for fire risk assessment, and vice versa.

[0122] The smoke concentration and temperature of the warehouse area under normal conditions (when there is no fire risk trend) within 60 seconds are collected in advance as the standard smoke concentration sequence and standard temperature sequence respectively.

[0123] The DTW (Dynamic Time Warping) algorithm is used to calculate the DTW distance between the historical smoke concentration series and the standard smoke concentration series at the current moment. The inverse of the DTW distance reflects the similarity between the historical smoke concentration series and the standard smoke concentration series. The higher the similarity, the closer the smoke concentration at the current moment is to the historical normal state, the less likely the smoke concentration at the current moment is to be abnormal, and the lower the potential value for risk assessment, and vice versa.

[0124] Similarly, the DTW algorithm is used to calculate the DTW distance between the historical temperature series at the current moment and the standard temperature series. The inverse of the DTW distance reflects the similarity between the historical temperature series and the standard temperature series. The higher the similarity, the closer the temperature at the current moment is to the historical normal state, the less likely the temperature at the current moment is to be abnormal, and the lower the potential value for risk assessment, and vice versa.

[0125] In summary, the comprehensive entropy and similarity are used to evaluate the comprehensive value index of smoke concentration and temperature at the current moment:

[0126]

[0127] In this formula, for The comprehensive value index of smoke concentration and temperature at the moment, for The information entropy value of the historical smoke concentration sequence at time for The similarity between the historical smoke concentration sequence and the standard smoke concentration sequence at the moment for The information entropy of the historical temperature sequence at the time, for The similarity between the historical temperature series and the standard temperature series at the moment.

[0128] In this formula, higher entropy values indicate greater data volatility, potentially representing a higher risk. In fire monitoring, higher entropy values typically indicate a significant difference between the current data and historical norms, potentially indicating a higher fire risk. Information entropy, as a fundamental metric, can provide an assessment of environmental data itself. The similarity metric assesses how closely the current smoke concentration and temperature match normal conditions. Higher similarity indicates that the current temperature and smoke concentration are close to normal, indicating a lower risk and lower value. Conversely, lower similarity indicates a significant difference between the current data and historical norms, increasing the risk and the likelihood of potential abnormal events, and the value index accordingly.

[0129] In the formula and When the data differs greatly from the standard state, the comprehensive value index increases, which directly reflects the abnormality and risk at the current moment. The product of information entropy further amplifies the comprehensive value index through the uncertainty of the data, reflecting the potential risk of higher uncertainty.

[0130] In summary, this formula dynamically calculates information entropy and similarity, allowing it to adjust its assessment value as environmental conditions change. Under normal circumstances, temperature and smoke concentration may be highly similar to standard conditions, resulting in low information entropy and a low value index. However, once a fire occurs, the volatility of temperature and smoke concentration increases, information entropy rises, and similarity to standard conditions decreases, leading to a higher value index, helping to promptly identify potential fire risks. This approach evaluates the potential value of current smoke concentration and temperature for fire risk assessment from multiple dimensions.

[0131] S4: Use the comprehensive risk index and comprehensive value index to dynamically adjust the collection frequency of the smoke concentration sensor and temperature sensor to achieve an adaptive data collection strategy.

[0132] In warehouse fire monitoring, the real-time and accurate sensor data is crucial for risk assessment. However, due to the complexity and uncertainty of environmental changes, a fixed collection frequency often cannot flexibly respond to scenarios with different risk levels.

[0133] By adaptively adjusting the data collection strategy, the data collection frequency can be dynamically optimized based on the current risk assessment results, thereby improving system efficiency and responsiveness. The key benefit of adaptive adjustment is that when the risk is low, the collection frequency is reduced to save energy and storage resources; when the risk is high, the collection frequency is increased to ensure timely acquisition of critical data, improving monitoring accuracy and responsiveness. This dynamic adjustment effectively responds to complex environmental changes, enhances the system's intelligence and flexibility, and provides more accurate data support for fire warnings.

[0134] Specifically, the maximum acquisition frequency and minimum acquisition frequency of the smoke concentration sensor are obtained based on the hardware parameters of the smoke concentration sensor and the temperature sensor. and , the maximum acquisition frequency and minimum acquisition frequency of the temperature sensor are and , the initial acquisition frequencies of the smoke concentration sensor and temperature sensor in step S1 are and .

[0135] For smoke concentration sensors, the acquisition frequency is dynamically adjusted based on the following formula:

[0136]

[0137] In the formula, The acquisition frequency after the smoke concentration sensor is adjusted. is the initial acquisition frequency of the smoke concentration sensor, for The comprehensive risk index of temperature and smoke concentration at the time, is a symbolic function, for The comprehensive value index of temperature and smoke density at the moment, for The theoretical middle value is 0.5 (empirical value).

[0138] In this formula, reflect The product of the comprehensive risk index and the comprehensive value index at the moment. The larger the product, the greater the risk and the higher the credibility. The more the smoke concentration sensor's acquisition frequency should be increased to capture high-value, high-risk smoke changes at a higher frequency; vice versa. When the two are closer to 1, the product is closer to 1, reaching the maximum value, indicating that the acquisition frequency needs to be increased to the greatest extent. If either is close to 0, it means that the risk is not high or the value is not great. The acquisition frequency can be appropriately reduced to save resources. The theoretical intermediate value of this product is Used to measure relative size, The middle value is a symmetrical middle point in terms of value. It is equal to a neutral value (reference value) of the product of risk and value index. It means that when the value of the product is close to When the risk and value index are in a balanced state, neither biased towards high risk nor low risk, the sensor should maintain the original acquisition frequency. When the product is less than When , it indicates that the fire risk is low and the credibility is poor, the collection frequency should be reduced to save system resources.

[0139] In this formula, reflects The difference between the theoretical intermediate value and Greater than , indicating that the risk factor is high and the value index is large, and the collection frequency of the smoke concentration sensor needs to be increased. If the value is greater than 0, the collection frequency of the smoke density sensor will be increased to better capture fire risks and provide timely warnings. The greater the difference, the greater the increase. equal , , indicating that the risk factor and value index are at a medium level. Keep the original acquisition frequency of the smoke density sensor. Less than , , indicating that the risk coefficient and value index are both low at this time, and the collection frequency of the smoke concentration sensor is reduced to save resources. The greater the difference, the greater the reduction.

[0140] Furthermore, we add boundary constraints: , to ensure that the collection frequency of the smoke concentration sensor does not exceed the normal range.

[0141] For temperature sensors, the acquisition frequency is dynamically adjusted based on the following formula:

[0142]

[0143] In the formula, is the acquisition frequency after the temperature sensor is adjusted, is the initial acquisition frequency of the temperature sensor, for The comprehensive risk index of temperature and smoke concentration at the time, is a symbolic function, for The comprehensive value index of temperature and smoke density at the moment, for The theoretical middle value is 0.5 (empirical value).

[0144] In this formula, This value reflects the product of the comprehensive risk index and the comprehensive value index at time t. The larger the product, the greater the risk and the higher the credibility. Therefore, the collection frequency should be increased to capture high-value, high-risk temperature changes. The reverse is also true. As both values approach 1, the product approaches 1, reaching its maximum value, indicating that the temperature sensor collection frequency should be increased to the maximum. If either value is close to 0, the risk is low or the value is low, and the temperature sensor collection frequency can be appropriately reduced to conserve resources.

[0145] In this formula, reflects The difference between the theoretical intermediate value and Greater than , indicating that the risk factor is high and the value index is large, and the acquisition frequency of the temperature sensor needs to be increased. If the value is greater than 0, the acquisition frequency of the temperature sensor will be increased to better capture fire risk data and issue early warnings. The greater the difference, the greater the increase. equal , indicating that the risk coefficient and value index are at a medium level. ,pass Keep the original sampling frequency of the temperature sensor. Less than , , indicating that the risk coefficient and value index are both low at this time, and the acquisition frequency of the temperature sensor is reduced to save resources. The greater the difference, the greater the reduction.

[0146] Furthermore, we add boundary constraints: , ensuring that the acquisition frequency of the temperature sensor does not exceed the normal range.

[0147] In summary, the acquisition frequency adjustment mechanism combines the comprehensive risk index and the comprehensive value index to dynamically adjust the initial acquisition frequency of the smoke concentration sensor and temperature sensor, demonstrating good scenario adaptability and responsiveness. When the system detects high risk and high data value, it automatically increases the sampling frequency, helping to capture key anomaly information in a timely manner. Conversely, when both risk and value are low, the sampling frequency is reduced to conserve resources. This method introduces a theoretical intermediate value as an adjustment reference, combined with sign functions and boundary control, to ensure clear adjustment logic, simple implementation, and stable frequency within a reasonable range. This method reflects the system's ability to dynamically balance energy consumption and risk perception, improving the intelligence level and practical application efficiency of the acquisition system.

[0148] The present invention also provides a data acquisition system based on Internet of Things devices, including: a temperature sensor and a smoke concentration sensor, respectively installed in the cargo area of the warehouse; a control unit, integrated with a computer program, which, when executed, responds to the data collected by the smoke concentration sensor and the temperature sensor and executes the data acquisition method described in steps S1 to S4 to implement an adaptive data acquisition strategy.

Claims

1. A data collection method based on IoT devices, characterized in that: include: Install smoke concentration sensors and temperature sensors in the cargo area of the warehouse; Taking any moment as the current moment, determine the smoke concentration risk index and temperature risk index at the current moment based on the respective deviation degrees of the smoke concentration and temperature at the current moment, as well as the respective change rates of the smoke concentration and temperature at the current moment; Determining the mutual information of the smoke concentration and temperature at the current moment based on the historical smoke concentration sequence and the historical temperature sequence at the current moment, and determining the comprehensive risk index of the smoke concentration and temperature at the current moment by combining the mutual information with the smoke concentration risk index and the temperature risk index; Determine the comprehensive value index of the current smoke concentration and temperature based on the uncertainty of the historical smoke concentration series and the historical temperature series and their similarity with the corresponding standard series; the uncertainty is determined based on the information entropy value; The comprehensive risk index and comprehensive value index are used to dynamically adjust the collection frequency of the smoke density sensor and the temperature sensor, including: ； In the formula, The adjusted acquisition frequency for the smoke density sensor or temperature sensor. is the initial acquisition frequency of the smoke density sensor or temperature sensor, for The comprehensive risk index of temperature and smoke concentration at the time, is a symbolic function, for The comprehensive value index of temperature and smoke density at the moment, for The theoretical median value of and It corresponds to the smoke concentration sensor or the temperature sensor at the same time to realize the adaptive data collection strategy; The combined risk index for smoke concentration and temperature includes: ； Where, for The comprehensive risk index of smoke concentration and temperature at the moment, for The smoke concentration risk index at the moment, for Temperature risk index at the moment, for The weight of the smoke concentration risk index at the moment, for The weight of the temperature risk index at the moment, is the normalization function, is the mutual information between smoke concentration and temperature at the current moment, for The information entropy value of the historical smoke concentration sequence at time for The information entropy value of the historical temperature series at the moment; The combined value index of smoke concentration and temperature includes: ； In the formula, for The comprehensive value index of smoke concentration and temperature at the moment, for The information entropy value of the historical smoke concentration sequence at time for The similarity between the historical smoke concentration sequence and the standard smoke concentration sequence at the moment for The information entropy of the historical temperature sequence at the time, for The similarity between the historical temperature series and the standard temperature series at the moment; The smoke concentration risk index and temperature risk index include: ； In the formula, for The risk index of smoke concentration or temperature at the moment, for The degree of deviation of smoke concentration or temperature at the moment, for The degree of deviation of smoke concentration or temperature at the moment, To take the absolute value sign, for The rate of change of smoke concentration or temperature at a given moment, is the median of the rate of change of the historical smoke concentration series or historical temperature series at time t; 、 、 、 and It corresponds to the smoke concentration or temperature at the same time.

2. The data collection method based on IoT devices according to claim 1, characterized in that: Dynamically adjusting the acquisition frequencies of the smoke concentration sensor and the temperature sensor also includes adding boundary limits to the adjusted acquisition frequencies of the smoke concentration sensor and the temperature sensor, including: Obtain the maximum acquisition frequency and minimum acquisition frequency of each according to the hardware parameters of the smoke density sensor and the temperature sensor; If the adjusted acquisition frequency of the smoke concentration sensor or the temperature sensor is less than the respective minimum acquisition frequency, the adjusted acquisition frequency is set to the respective minimum acquisition frequency; If the adjusted collection frequency of the smoke concentration sensor or the temperature sensor is greater than the respective minimum collection frequency, the adjusted collection frequency is set to the respective maximum collection frequency.

3. The data collection method based on IoT devices according to claim 1 is characterized in that smoke The weights of the concentration risk index and the temperature risk index are determined based on the following: Calculate the cumulative value of the coefficient of variation of the historical smoke concentration series at the current moment and the coefficient of variation of the historical temperature series; use the ratio of the coefficient of variation of the historical smoke concentration series at the current moment to the cumulative value as the weight of the smoke concentration risk index at the current moment; The ratio of the coefficient of variation of the historical temperature series at the current moment to the accumulated value is used as the weight of the temperature risk index at the current moment.

4. The data collection method based on IoT devices according to claim 1, characterized in that: The historical smoke concentration series and historical temperature series are determined as follows: All historical smoke concentrations and all temperatures in a preset time period before the current moment are respectively used to form a historical smoke concentration sequence and a historical temperature sequence at the current moment.

5. The data collection method based on IoT devices according to claim 4, characterized in that: The standard smoke concentration series and standard temperature series are determined based on the following methods: The smoke concentration and temperature of the cargo area under normal conditions within a preset time period are collected in advance to form a standard smoke concentration sequence and a standard temperature sequence respectively; Furthermore, the preset time period is the same as the length of the preset time period when obtaining the historical smoke concentration sequence and the historical temperature sequence.

6. A data acquisition system based on IoT devices, characterized in that: include: Temperature sensors and smoke concentration sensors are installed in the cargo area of the warehouse; The control unit is integrated with a computer program, which, when executed, responds to data collected by the smoke concentration sensor and the temperature sensor and executes the data collection method according to any one of claims 1 to 5 to implement an adaptive data collection strategy.

Images