A method and system for early warning of ventilation equipment in a smart laboratory

Abstract

The present application relates to the field of ventilation equipment early warning, and more particularly to a kind of ventilation equipment early warning method and system of wisdom laboratory, method includes: with target time as center to construct window, obtain the multidimensional state parameter sequence of ventilation equipment in window;Maximum cross-correlation coefficient of target time is calculated, and optimal lag is obtained, and final lag coefficient is obtained by weighting optimal lag, and cross-correlation difference significance is obtained according to the maximum cross-correlation coefficient of each sampling time in window;State parameter sequence of different dimensions in window is aligned based on final lag coefficient, the correlation of state parameter sequence of any two dimensions is obtained, the risk value of target time is calculated, and the risk correction value is obtained by correcting risk value.The technical scheme of the present application can improve the accuracy of ventilation equipment early warning and reduce the impact of hysteresis on early warning.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of ventilation equipment early warning. In particular, it relates to a ventilation equipment early warning method and system for a smart laboratory. BACKGROUND

[0002] In the construction of modern smart laboratories, environmental control is one of the cores of laboratory safety and efficiency. The equipment and experimental processes involved in the laboratory may produce various harmful gases, heat, dust, and volatile substances, which need to be regulated through precise ventilation systems. However, traditional ventilation control methods mainly rely on manual monitoring and timed switches, which are prone to not responding to environmental changes in a timely manner, failing to accurately adapt to various complex experimental conditions, and not being able to achieve precise early warning and timely response. SUMMARY

[0003] To solve the above technical problems, the present application provides solutions in the following aspects.

[0004] In a first aspect, a ventilation equipment early warning method for a smart laboratory includes: constructing a window centered on a target time, the target time being any sampling time, obtaining a multi-dimensional state parameter sequence of the ventilation equipment in the window; calculating the maximum cross-correlation coefficient of the target time and obtaining the optimal lag, weighting the optimal lag to obtain the final lag coefficient, and obtaining the cross-correlation difference significance based on the maximum cross-correlation coefficient of each sampling time in the window; aligning the state parameter sequences of different dimensions in the window based on the final lag coefficient, obtaining the correlation of the state parameter sequences of any two dimensions, calculating the risk value of the target time based on the minimum value of the correlation, the cross-correlation difference significance, the maximum cross-correlation coefficient and the final lag coefficient, correcting the risk value to obtain a risk correction value, and generating and sending an alarm signal in response to the risk correction value being greater than a preset risk threshold.

[0005] Preferably, the calculation of the maximum cross-correlation coefficient of the target time includes: presetting an initial lag value and a change range of the initial lag value for the target time, using a piecewise convolution method to calculate the cross-correlation coefficient of any two dimensions under the same initial lag value, and calculating the cross-correlation coefficient cumulative value; traversing to obtain the cross-correlation coefficient cumulative value corresponding to each initial lag value, and selecting the maximum value of the cross-correlation coefficient cumulative value as the maximum cross-correlation coefficient of the target time.

[0006] Preferably, the optimal lag includes: taking the initial lag value corresponding to the maximum cross-correlation coefficient as the optimal lag.

[0007] Preferably, the weighting of the optimal lag to obtain the final lag coefficient comprises: taking the ratio of the maximum cross-correlation coefficient of the target moment to the cumulative value of the maximum cross-correlation coefficients of the window as the weight of the target moment, calculating the first product of the weight of the target moment and the optimal lag, and taking the cumulative value of all first products in the window as the final lag coefficient.

[0008] Preferably, the method for obtaining the cross-correlation significance comprises: rank sum test, sign test or Ridit test.

[0009] Preferably, the calculation of the risk value comprises: taking the difference between 1 and the minimum value of the correlation as the first term; taking the sum of the cross-correlation significance and the preset hyperparameter through the mapping result of the logarithmic function as the second term; calculating the standard deviation of the maximum cross-correlation coefficient of all sampling moments in the window, and calculating the first ratio of the standard deviation and the final lag coefficient, taking the difference between 1 and the first ratio as the alignment consistency; and taking the product of the first term, the second term and the alignment consistency as the risk value.

[0010] Preferably, the correction of the risk value to obtain the risk correction value comprises: taking the previous sampling moment adjacent to the target moment as the control moment, obtaining the risk correction value of the control moment; taking the absolute value of the final lag coefficient through the negative correlation mapping result as the lag weight, calculating the second product of the lag weight and the risk value of the target moment; calculating the difference between 1 and the lag weight as the weight difference, calculating the third product of the weight difference and the risk correction value of the control moment; and taking the sum of the second product and the third product as the risk correction value of the target moment.

[0011] The second aspect is a ventilation equipment early warning system of a smart laboratory, comprising: a processor and a memory, the memory stores computer program instructions, when the computer program instructions are executed by the processor, a smart laboratory ventilation equipment early warning method is realized.

[0012] The present application has the following effects:

[0013] The present application constructs a window centered on the target moment, analyzes the multi-dimensional state parameters of the ventilation equipment in detail, accurately describes the equipment running state by using the maximum cross-correlation coefficient and the lag coefficient, quantitatively evaluates the correlation and significance between different state parameters, can effectively identify potential risks, and thus early warning. The present application not only ensures the accuracy and reliability of real-time monitoring, but also provides strong data support for subsequent ventilation equipment maintenance and fault prediction, helps to optimize the performance of the ventilation equipment, and improves the energy efficiency and safety of the overall system. In risk management, considering the lag effect makes the decision more in line with the actual situation, and finally realizes more scientific and efficient ventilation equipment management and risk control, thereby laying a foundation for protecting the air quality and personnel safety in the space. BRIEF DESCRIPTION OF DRAWINGS

[0014] Figure 1 is a flow chart of a ventilation equipment early warning method of a smart laboratory according to an embodiment of the present application. DETAILED DESCRIPTION

[0015] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some embodiments of the present application, but not all the embodiments of the present application.

[0016] The specific embodiments of the present application will be described in detail below with reference to the drawings.

[0017] Referring to Figure 1 A ventilation equipment early warning method of a smart laboratory includes steps S1-S3, which are specifically as follows:

[0018] S1: A window is constructed with a target time as the center, the target time is any sampling time, and a multi-dimensional state parameter sequence of the ventilation equipment in the window is obtained.

[0019] In an embodiment, a time window is constructed with a target time as the center, and the target time can be any sampling time. After the target time is determined, a window containing multiple time points is established around the time, so as to deeply analyze the performance of the ventilation equipment at this specific time. In this window, a multi-dimensional state parameter sequence of the ventilation equipment is collected.

[0020] Among the multi-dimensional parameters, the valve opening degree is an important indicator, reflecting the state of airflow regulation in the ventilation system. By monitoring the opening degree of the valve, the ventilation volume change and operation efficiency of the system can be understood, so as to determine whether the equipment is running normally.

[0021] Air volume directly affects the effect of ventilation and the law of air flow. Proper air volume not only ensures the air quality in the space, but also reduces energy consumption and improves the energy efficiency of the system. Therefore, the monitoring of air volume provides necessary data support for optimizing the performance of the ventilation equipment.

[0022] Pressure difference is also a key multi-dimensional parameter. This indicator is usually used to evaluate the difference between the internal and external air pressure in the ventilation system, which affects the direction and rate of air flow. By analyzing the pressure difference, the working state of the ventilation equipment can be effectively judged, and potential faults or abnormal phenomena can be found in time.

[0023] When constructing this window centered on the target time, all data within the window can be integrated and analyzed. By cross-comparing multi-dimensional status parameters such as valve opening, airflow, and differential pressure, the operating status of ventilation equipment can be comprehensively understood and evaluated. Such analysis not only helps in real-time monitoring of equipment performance but also provides a reliable basis for subsequent fault prediction, performance evaluation, and optimized management.

[0024] Furthermore, combining data from the target time with nearby time points allows for the observation of fluctuations in the performance of ventilation equipment across different time periods, thus supporting time series analysis. This method can more clearly identify trends and anomalies in daily operation, enabling timely adjustments and optimization decisions.

[0025] For example, the window length is 11, that is, with the target time as the center, there are 5 sampling times on the left side of the center and 5 sampling times on the right side of the center. The specific window length can be set by those skilled in the art.

[0026] S2: Calculate the maximum cross-correlation coefficient at the target time and obtain the optimal lag. Weight the optimal lag to obtain the final lag coefficient. Obtain the significance of cross-correlation difference based on the maximum cross-correlation coefficient at each sampling time in the window.

[0027] In one embodiment, for a target time, a preset lag initial value and a range of variation for the lag initial value are defined. Under the same lag initial value, the cross-correlation coefficients of any two dimensions are calculated using the piecewise convolution method, and the cumulative value of the cross-correlation coefficients is calculated. The cumulative value of the cross-correlation coefficients corresponding to each lag initial value is obtained by traversal, and the maximum value of the cumulative value of the cross-correlation coefficients is selected as the maximum cross-correlation coefficient at the target time.

[0028] For example, the target time is the 10th second, and the range of the initial lag value is between -2 and +2. For any initial lag value of +1, i.e., a delay of 1 second, according to step S1, the multiple dimensions are valve opening, air volume, and pressure difference. It is necessary to calculate the cross-correlation coefficients between valve opening and air volume, valve opening and pressure difference, and air volume and pressure difference for a delay of 1 second, and then sum these three cross-correlation coefficients to obtain the cumulative cross-correlation coefficient value for a delay of 1 second. The same method is used to obtain the cumulative cross-correlation coefficient values ​​for the target time at different initial lag values. The maximum value among the cumulative cross-correlation coefficient values ​​for different initial lag values ​​is selected as the maximum cross-correlation coefficient for the target time.

[0029] The initial lag value corresponding to the largest cross-correlation coefficient is taken as the optimal lag.

[0030] The final lag coefficient is obtained by weighting the optimal lag, including: Ratioing the maximum cross-correlation coefficient at the target time to the sum of the maximum cross-correlation coefficients calculated within the window, thus determining the weight of the target time. This weight reflects the relative importance of the target time in the overall analysis. This weight is then multiplied by the optimal lag to obtain the first product. All calculated first products within the window are then summed to obtain the final lag coefficient. This lag coefficient comprehensively considers the impact of the target time and the overall correlation of the window data, providing a more comprehensive reflection of the ventilation equipment's state changes over different time periods.

[0031] When analyzing the maximum cross-correlation coefficients at each sampling time point within the analysis window, a significance test is used to assess the significant differences between these cross-correlation coefficients. By implementing different statistical tests, such as the rank-sum test, the sign test, or the Ridit test, it is possible to accurately determine whether there are statistically significant differences in the cross-correlation coefficients at each time point. These tests each have unique application scenarios and advantages, and can identify whether fluctuations around the expected value are merely due to randomness or reflect deeper changes in correlation. Thus, obtaining numerical results for the cross-correlation coefficients through significance testing provides solid data support and theoretical basis for subsequent decision-making and optimization.

[0032] S3: Align the state parameter sequences of different dimensions in the window based on the final lag coefficient, obtain the correlation between any two state parameter sequences, calculate the risk value at the target time based on the minimum correlation, the significance of cross-correlation difference, the maximum cross-correlation coefficient and the final lag coefficient, correct the risk value to obtain the risk correction value, and generate and send an alarm signal when the risk correction value is greater than the preset risk threshold.

[0033] In one embodiment, the final hysteresis coefficient is used to align the state parameter sequences of different dimensions within a window to ensure accurate temporal correspondence. This alignment allows for effective comparison of state parameter sequences of any two dimensions, enabling the calculation of their correlation. This process allows for synchronous analysis of data from different dimensions, avoiding errors caused by time differences and providing a clearer and more realistic understanding of the relationships between parameters. After alignment, the dynamic interactions and correlations between various dimensions can be revealed, providing crucial information for performance evaluation, fault diagnosis, or optimized management of ventilation equipment.

[0034] Based on the aligned multi-dimensional state parameter sequence, the correlation between any two dimensions can be calculated, and the minimum value of all correlations can be obtained.

[0035] The risk value at the target time is calculated based on the minimum correlation, the significance of cross-correlation differences, the maximum cross-correlation coefficient, and the final lag coefficient. The risk value satisfies the following relationship:

[0036] , Indicates the target time The risk value, Represents the minimum value of correlation. Indicates the target time The significance of the cross-correlation difference This represents the standard deviation of the maximum cross-correlation coefficient across all sampling times within the window. Indicates the final lag factor. Represents the logarithmic function. The purpose of representing hyperparameters is to prevent The value can be 0. For example, The value is 0.0001.

[0037] When alignment is good, a lower correlation value and a higher significance value indicate a higher hazard value. Constructing a hazard value centered on the target time point can effectively assess the correlation and significance between variables across different time windows. Specifically, a lower correlation value indicates a weaker dependency between variables, while a higher significance value signifies a significant statistical relationship. In this case, the hazard value will increase due to increased potential uncertainty.

[0038] The risk value is corrected to obtain a risk correction value, including: taking the previous sampling time adjacent to the target time as the control time and obtaining the risk value of the control time; taking the absolute value of the final lag coefficient as the result of negative correlation mapping as the lag weight, and calculating the second product of the lag weight and the risk value of the target time; calculating the difference between 1 and the lag weight as the weight difference, and calculating the third product of the weight difference and the risk value of the control time; and taking the sum of the second product and the third product as the risk correction value of the target time.

[0039] The risk adjustment value satisfies the following relationship:

[0040] , Indicates the target time The risk correction value, Indicates the final lag factor. Indicates the target time The risk value, Indicates the control time The risk correction value, This represents an exponential function.

[0041] When the final lag coefficient When the value is larger, the exponential function The value will become smaller and smaller, which makes... The closer it is to 1, the greater the... This means the memory effect lasts longer, thus amplifying the delay effect. This effect directly amplifies the risk value because the system's "memory" of past states becomes more significant and persistent when responding to changes, potentially leading to a slower response to the current state. Specifically, a longer memory period may prevent the system from adjusting in time to mitigate risk when faced with sudden events or changes, potentially leading to risk accumulation and exacerbation. Therefore, in risk management, special attention needs to be paid to larger [risks / problems]. To mitigate the potential risks, it is crucial to strike a reasonable balance between latency and response speed to avoid the negative impact of excessively long memory cycles on system stability.

[0042] When analyzing the risk values ​​of ventilation equipment, the impact of lag issues needs to be considered. When the lag effect is significant, the risk warning system for ventilation equipment becomes more reliant on the cumulative error value trending upwards. This means that when facing increasingly severe risk situations, the system needs to not only focus on current real-time data but also give higher weight to historical data to correct the risk value calculation. In this way, the influence of historical data will increase as the risk intensifies, thus more accurately reflecting the true risk situation of the equipment. Ultimately, the corrected risk value will be able to more effectively guide risk management and decision-making, helping to take timely measures to reduce potential hazards.

[0043] The risk correction value is compared with the preset risk threshold. If the risk correction value is greater than the preset risk threshold, an alarm signal is generated and sent.

[0044] The system includes a processor and a memory, the memory storing computer program instructions, which, when executed by the processor, implement a smart laboratory ventilation equipment early warning method according to the first aspect of the present invention.

[0045] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

[0046] It should be noted that those skilled in the art can make various modifications and improvements without departing from the inventive concept, and these all fall within the scope of protection of this invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for early warning of ventilation equipment in a smart laboratory, characterized in that, include: A window is constructed with the target time as the center, where the target time is any sampling time, to obtain the multi-dimensional state parameter sequence of the ventilation equipment within the window; Calculate the maximum cross-correlation coefficient at the target time and obtain the optimal lag. Weight the optimal lag to obtain the final lag coefficient. Obtain the significance of cross-correlation difference based on the maximum cross-correlation coefficient at each sampling time in the window. Align the state parameter sequences of different dimensions in the window based on the final lag coefficient, obtain the correlation between any two state parameter sequences, calculate the risk value at the target time based on the minimum value of the correlation, the significance of the cross-correlation difference, the maximum cross-correlation coefficient and the final lag coefficient, correct the risk value to obtain the risk correction value, and generate and send an alarm signal when the risk correction value is greater than the preset risk threshold. The maximum cross-correlation coefficient calculated at the target time includes: For the target time, a preset initial lag value and a range of variation for the initial lag value are defined. Under the same initial lag value, the cross-correlation coefficient of any two dimensions is calculated using the piecewise convolution method, and the cumulative value of the cross-correlation coefficient is calculated. Iterate through each lag initial value to obtain the cumulative cross-correlation coefficient value, and select the maximum value of the cumulative cross-correlation coefficient value as the maximum cross-correlation coefficient at the target time. The process of obtaining the optimal lag includes: The initial lag value corresponding to the maximum cross-correlation coefficient is taken as the optimal lag; The process of weighting the optimal lag to obtain the final lag coefficient includes: The ratio of the maximum cross-correlation coefficient at the target time to the sum of the maximum cross-correlation coefficients within the window is used as the weight of the target time. The first product of the weight of the target time and the optimal lag is calculated, and the sum of all first products within the window is used as the final lag coefficient.

2. The early warning method for ventilation equipment in a smart laboratory according to claim 1, characterized in that, Methods for obtaining the significance of the cross-correlation difference include: Rank sum test, sign test, or Ridit test.

3. The early warning method for ventilation equipment in a smart laboratory according to claim 1, characterized in that, Calculating the risk value includes: The difference between 1 and the minimum value of the correlation is taken as the first term; The result of mapping the significance of cross-correlation differences to the sum of preset hyperparameters using a logarithmic function is taken as the second term; Calculate the standard deviation of the maximum cross-correlation coefficient at all sampling times within the window, and calculate the first ratio of the standard deviation to the final lag coefficient. Use the difference between 1 and the first ratio as the alignment consistency. The product of the first term, the second term, and the alignment consistency is used as the risk value.

4. The early warning method for ventilation equipment in a smart laboratory according to claim 1, characterized in that, The process of correcting the risk value to obtain a risk correction value includes: The previous sampling time adjacent to the target time is used as the control time to obtain the risk correction value of the control time; The absolute value of the final lag coefficient is used as the result of negative correlation mapping as the lag weight, and the second product of the lag weight and the risk value at the target time is calculated. Calculate the difference between 1 and the lag weight as the weight difference, and calculate the third product of the weight difference and the risk correction value at the control time. The sum of the second and third products is used as the risk correction value at the target time.

5. A smart laboratory ventilation equipment early warning system, characterized in that, include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement a ventilation equipment early warning method for a smart laboratory according to any one of claims 1-4.

Images