Intelligent monitoring and early warning system for food production environment

Abstract

The present application relates to the technical field of food production, in particular to an intelligent monitoring and early warning system for food production environment, which comprises an environmental parameter monitoring module, an active feature identification module, an induced event analysis module, a dynamic window construction module and an early warning linkage output module. The present application integrates temperature and humidity fluctuations and infrared energy changes, establishes a multi-parameter joint analysis mechanism, takes the dynamic sampling period and fluorescence response trend as the core of the microbial growth state, constructs a closed-loop process from environmental change to risk identification through continuous tracking of feature nodes and environmental disturbance response, realizes real-time mapping of wet and hot parameter changes and active signal cycle characteristics, realizes synchronous linkage of production site control records and microbial concentration growth timing, realizes adaptive adjustment of risk window through dynamic window expansion, and directly feeds back the early warning signal according to the microbial activity stage and the on-site environment, forming the overall ability of early capture, dynamic tracking and accurate response of pollution risk.

Description

Technical Field

[0001] This invention relates to the field of food production technology, and in particular to an intelligent monitoring and early warning system for the food production environment. Background Technology

[0002] Food production involves environmental control and management of food processing, packaging, storage, transportation, and related facilities. Monitoring and controlling the production environment is a crucial aspect of ensuring food safety and quality, requiring real-time monitoring of indicators such as air temperature and humidity, harmful gases, and microbial concentrations to reduce contamination risks and meet regulatory standards. Traditional intelligent monitoring and early warning systems for food production environments refer to systems that collect and monitor environmental parameters within food processing or packaging areas using installed equipment and technologies, and issue early warning signals when detected data exceeds set thresholds.

[0003] Existing technologies rely solely on static settings and single thresholds for environmental monitoring, lacking the ability to identify dynamic interactions among multiple parameters. The system struggles to continuously track changes in temperature, humidity, air disturbance, and microbial concentration. Judgments of environmental anomalies depend on single-cycle data mutations, easily overlooking potential risk phases. The determination of trends in microbial activity growth is delayed, and there is a lack of support for analyzing the linkage between environmental control measures in production sites and microbial dynamics. As a result, key stages between environmental changes and microbial risks are not identified, early signals of pollution spread cannot be captured in a timely manner, and there are problems of untimely response and inaccurate judgment between the early warning stage and the actual risk state. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and propose an intelligent monitoring and early warning system for the food production environment.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: an intelligent monitoring and early warning system for the food production environment, the system comprising:

[0006] The environmental parameter monitoring module is based in the food production workshop. It analyzes real-time temperature and humidity data streams, screens temperature and humidity fluctuations caused by air flow, judges changes in infrared spectral moisture energy, dynamically adjusts the sampling cycle, and obtains a set of environmental change parameters.

[0007] The activity feature identification module analyzes fluorescence response signals based on the environmental change parameter set, determines the rate of change of microbial reaction curves, compares the consistency of activity cycles and trends, identifies concentration growth samples, and obtains growth feature node groups.

[0008] Based on the growth feature node group, the triggering event analysis module determines the synchronization of the feature node with the current temperature and humidity on the time axis, identifies the parameter change cycle in the adjustment log, locates the synchronous segment of the damp heat response, and obtains the linkage response interval.

[0009] The dynamic window construction module analyzes the corresponding bacterial incubation cycle based on the linkage response interval, identifies the increasing samples in the continuous monitoring cycle, judges the rise rate characteristics, adjusts the sampling frequency, and obtains a dynamic growth window group.

[0010] The early warning linkage output module, based on the dynamically growing window group, judges the microbial concentration fluctuation in the current monitoring period, compares the detection results with the linkage triggering standard, and obtains the early warning linkage triggering signal.

[0011] The present invention is improved in that the environmental change parameter set includes air state characteristic quantities, sensor monitoring area distribution and abnormal fluctuation records; the growth characteristic node group includes the bacterial community growth starting point, concentration change trend point and activity peak identifier; the linkage response interval includes linkage discrimination index, synchronous response period and control boundary attribute; the dynamic growth window group includes growth window number, incubation characteristic differentiation and data time series label; and the early warning linkage trigger signal includes risk response identifier, linkage execution parameter and record archiving identifier.

[0012] The present invention is improved in that the environmental parameter monitoring module includes:

[0013] The data stream receiving submodule is based on the food production workshop. It analyzes the time series data generated by the temperature and humidity sensors, classifies each sampling point according to its spatial location and collection time, compares the temperature and humidity changes of each sampling point within the same period, and judges the data continuity and the integrity of the sampling point distribution to obtain the periodic monitoring data group.

[0014] The fluctuation identification submodule, based on the periodic monitoring data set, determines the continuous periodic temperature and humidity changes at each sampling point, identifies segments with changes, compares the characteristics of air disturbance parameter changes within each segment, marks the time periods of abnormal changes, and obtains an air disturbance feature set.

[0015] The energy trend determination submodule, based on the air disturbance feature set, simultaneously collects energy absorption data of multiple feature bands, determines the changing trend of each moisture channel within the monitoring period, adjusts the sampling interval configuration, and obtains the environmental change parameter set.

[0016] The present invention is improved in that the activity feature recognition module includes:

[0017] The fluorescence signal detection submodule analyzes the fluorescence response signal detected by the air sampling device based on the environmental change parameter set, organizes the signal data sequentially according to the sampling period, compares the rate of signal change in each period, determines the position where the signal curve begins to rise, and obtains the periodic fluorescence response sequence.

[0018] The activity change recognition submodule determines the change trend of each periodic signal based on the periodic fluorescence response sequence, calculates the signal growth rate within consecutive periods, compares the correspondence between the cell activity period parameters and the current signal trend, and obtains the activity period matching sequence.

[0019] The growth node extraction submodule, based on the active cycle matching sequence, filters out cycles with continuously increasing signals, determines the continuity of microbial concentration changes within the cycle, classifies and marks time points with change characteristics, and obtains growth characteristic node groups.

[0020] The present invention is improved in that the induced event analysis module includes:

[0021] The node synchronization discrimination submodule, based on the growth feature node group, determines the synchronization status of the time position of each feature node with the temperature and humidity changes within the corresponding sampling period, compares the temporal correlation between each node and the changes in environmental parameters, identifies the combination of nodes and environmental data with a trend correlation coefficient greater than a preset threshold, and obtains the node-environment temporal matching set.

[0022] The log period extraction submodule, based on the node environment time series matching set, determines the environmental adjustment equipment parameter change behavior in the workshop adjustment log within the associated time period, compares the equipment adjustment record with the operation activity of the node matching time period, identifies the log segment with operation change, and obtains the parameter disturbance segment set.

[0023] The response segment positioning submodule calculates the time synchronization length, temperature and humidity parameter fluctuation amplitude, and number of feature nodes between each segment based on the parameter disturbance segment set, using the following formula:

[0024] ;

[0025] Obtain the linkage strength factor of the response section By filtering out sections with high linkage intensity, the linkage response interval is obtained, among which... Indicates the total number of overlapping segments. Indicates the first Time synchronization length of each segment Indicates the first The normalized index of the temperature and humidity parameter fluctuation range of each section Indicates the first The number of matching feature nodes in each segment.

[0026] The present invention is improved in that the dynamic window construction module includes:

[0027] The incubation cycle extraction submodule analyzes the temperature and humidity sequences based on the linkage response interval, compares the combined temperature and humidity changes of each monitoring segment, determines whether the trend of the monitoring data changes conforms to the continuous characteristics of the incubation period of the strain, identifies the corresponding data segment, and obtains the microbial incubation cycle parameters.

[0028] The incremental sequence identification submodule compares the fluorescence intensity sequences of each monitoring period based on the microbial incubation cycle parameters, calculates the fluorescence change amplitude within the continuous period, identifies the data segment with continuously increasing change amplitude, determines the sample sequence of the continuously increasing segment, and obtains the concentration-increasing sample sequence.

[0029] The trend feature summarization submodule calculates the concentration shift between the endpoint of each segment and the previous period based on the concentration-increasing sample sequence, optimizes the joint processing of concentration change and fluorescence rate, obtains the dynamic growth window identifier value, judges the strength of trend feature summarization, adjusts the sampling interval, and obtains the dynamic growth window group.

[0030] The present invention is improved in that the early warning linkage output module includes:

[0031] The concentration fluctuation judgment submodule analyzes the changes in microbial concentration in each monitoring period within the dynamic growth window group based on the dynamic growth window group, judges the rise, fall or stagnation of concentration in each period, compares the trend of concentration changes in each stage in the continuous period, classifies the fluctuation state, and obtains the microbial concentration fluctuation state.

[0032] The linkage condition matching submodule compares the microbial concentration fluctuation state with the corresponding relationship of the linkage triggering standard within the monitoring period to determine whether the parameter combination within the period meets the triggering requirements and obtains linkage judgment matching information.

[0033] Based on the linkage determination and matching information, the signal archiving and output submodule analyzes temperature and humidity data, microbial concentration change nodes and equipment control commands, determines the start-up sequence of the audible and visual warning device and the air circulation device, archives the current cycle monitoring and control data information, and obtains the early warning linkage trigger signal.

[0034] The present invention is improved in that the data stream refers to a continuous time series of raw environmental data such as temperature and humidity continuously collected and transmitted by the temperature and humidity sensor, and the fluorescence response signal refers to the fluorescence intensity signal generated by microorganisms or other biological components under an excitation light source.

[0035] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0036] In this invention, by linking and integrating temperature and humidity fluctuations with infrared energy changes, a multi-parameter joint analysis mechanism is established. The microbial growth status is based on the dynamic sampling cycle and fluorescence response trend. Through continuous tracking of feature nodes and response to environmental disturbances, a closed-loop process from environmental change to risk identification is constructed. The changes in humidity and heat parameters are mapped in real time with the periodic characteristics of the activity signal. The control records at the production site are synchronized with the microbial concentration growth sequence. The dynamic window expansion enables adaptive adjustment of the risk window. The early warning signal is directly fed back based on the microbial activity stage and the on-site environment, forming an overall capability for early capture, dynamic tracking and precise response to pollution risks. Attached Figure Description

[0037] Figure 1 This is a system flowchart of the present invention;

[0038] Figure 2 This is a flowchart of the environmental parameter monitoring module in this invention;

[0039] Figure 3 This is a flowchart of the activity feature recognition module in this invention;

[0040] Figure 4 This is a flowchart of the event induction analysis module in this invention;

[0041] Figure 5 This is a flowchart of the dynamic window construction module in this invention;

[0042] Figure 6 This is a flowchart of the early warning linkage output module in this invention. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0044] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0045] All user-related information involved in this invention (including but not limited to biometric information, identity verification information, behavioral data, device information, and other data that can be used for identity verification and personalized services) is collected and processed with the user's full knowledge and voluntary consent. The collection, storage, and use of all information strictly comply with applicable national and regional laws and regulations, and meet relevant data protection standards and policy requirements. The use of data is limited to purposes necessary for providing the technical services of this invention, and reasonable technical and management measures will be taken to ensure the security and confidentiality of users' personal information in terms of information protection and privacy.

[0046] Example

[0047] Please see Figure 1 This invention provides a technical solution: an intelligent monitoring and early warning system for the food production environment, comprising:

[0048] The environmental parameter monitoring module is based in a food production workshop. It analyzes the data stream collected in real time by temperature and humidity sensors, compares the temperature and humidity change ranges at each sampling point, identifies the fluctuation parameters under the influence of air flow, judges the energy change trend of each moisture feature during infrared spectral scanning, adjusts the sensor sampling cycle, optimizes the anomaly identification sensitivity, and obtains a set of environmental change parameters.

[0049] The activity feature recognition module analyzes the fluorescence response signal detected by the air sampling device based on the environmental change parameter set, determines the rate of change of the microbial reaction curve in each sampling period, compares the consistency of the trend of the cell activity cycle parameter with the current data stage, identifies the sample sequence corresponding to the rapid increase of microbial concentration, and marks the feature nodes in the sampling period separately to obtain the growth feature node group.

[0050] The triggered event analysis module is based on the growth feature node group, judges the synchronization relationship between the feature node and the current temperature and humidity parameters on the time axis, analyzes the fluctuation of humidity and heat, identifies the period of frequent parameter fluctuation in the workshop adjustment log, compares the correlation of temperature and humidity setting changes within different time periods, locates the actual humidity and heat response synchronization segment, and obtains the linkage response interval.

[0051] The dynamic window construction module is based on the linkage response interval, analyzes the incubation cycle parameters of the microbial species corresponding to each humid and hot zone, identifies the sample sequence with increasing concentration within the continuous monitoring cycle of the microbial sampling chamber, judges the rising rate characteristics of the collected data segment, summarizes the trend of the final concentration change of each data stage, and adjusts the sampling frequency to obtain the dynamic growth window group.

[0052] The early warning linkage output module is based on a dynamically growing window group to determine the fluctuation state of microbial concentration in the current monitoring period, compare the periodic detection results with the linkage triggering standard, analyze the linkage timing of the audible and visual warning device and the air circulation device, archive the temperature and humidity monitoring data, bacterial node information and control signals of the current period, and obtain the early warning linkage triggering signal.

[0053] The environmental change parameter set includes air state characteristics, sensor monitoring area distribution, and abnormal fluctuation records. The growth characteristic node group includes the bacterial community growth starting point, concentration change trend point, and activity peak identifier. The linkage response interval includes linkage discrimination indicators, synchronous response period, and control boundary attributes. The dynamic growth window group includes growth window number, incubation characteristic differentiation, and data time series label. The early warning linkage trigger signal includes risk response identifier, linkage execution parameters, and record archiving identifier.

[0054] In the environmental parameter monitoring module, the data stream refers to the continuous time series of raw environmental data such as temperature and humidity continuously collected and transmitted by temperature and humidity sensors, which is the basic raw information for dynamic monitoring; the sampling point refers to the temperature and humidity monitoring points arranged at different spatial locations or time points in the food production workshop, with each sampling point representing a specific collection location or sampling time; the fluctuation parameter refers to the amplitude or frequency of changes in environmental data such as temperature and humidity within a short period of time, mainly used to identify abnormal environmental disturbances; infrared spectral scanning refers to the periodic scanning of the air using infrared sensing equipment to obtain energy absorption characteristic data of substances such as moisture in the air under different wavelengths; moisture characteristics refer to the specific absorption peaks exhibited by moisture (such as steam or tiny water droplets) in the infrared spectrum, which is the basis for determining humidity changes; the energy change trend refers to the rise and fall of energy in different moisture characteristic bands of the infrared spectrum as the sampling time progresses, used to assist in judging environmental changes; and the anomaly identification sensitivity refers to the sensitivity of perception and response to sudden changes in temperature, humidity, or infrared energy, which is achieved by adjusting parameters such as the sampling period.

[0055] In the activity feature identification module, air sampling equipment refers to instruments used to collect workshop air samples and transport them to subsequent analysis devices (such as fluorescence detectors), including components such as pumps and filters; fluorescence response signal refers to the fluorescence intensity signal produced by microorganisms or other biological components under a specific excitation light source, reflecting changes in the concentration of active microorganisms in the air; microbial reaction curve refers to the trajectory of microbial activity changes plotted with time or period as the horizontal axis and fluorescence or other activity detection signals as the vertical axis; cell activity cycle parameter refers to the typical time characteristics required for different microorganisms (species) to reproduce, grow, and stabilize in a suitable environment, serving as a reference for identifying growth nodes; current data stage refers to the current real-time sampling or monitoring time interval, corresponding to the latest batch of environmental and microbial detection data; sample sequence refers to the set of air samples and detection results obtained within multiple consecutive sampling periods; feature node refers to the moment in the sample sequence that represents a sharp change in microbial concentration or the reaching of a critical threshold, which is the focus of potential risk identification.

[0056] In the triggering event analysis module, synchronization relationship refers to the sequential or synergistic occurrence of microbial activity change nodes and environmental parameter (temperature, humidity, etc.) changes on the time axis; humidity and heat change fluctuations refer to the fluctuation phenomenon exhibited when environmental temperature and humidity change in tandem, especially the synchronous rapid rise and fall at certain stages; workshop adjustment log refers to the time record file of parameter adjustments, switching on and off, and setting changes of all environmental control equipment (such as air conditioners, humidifiers, etc.) in the workshop; the period of frequent parameter fluctuations refers to the monitoring cycle in which control parameters such as temperature and humidity are adjusted or changed multiple times within a certain period of time, which is an abnormally active stage of system monitoring; correlation refers to the correspondence between microbial growth nodes and temperature and humidity changes in statistics or events, used to infer the triggering mechanism; humidity and heat response synchronization segment refers to the specific monitoring interval in which environmental temperature and humidity changes and microbial activity events are highly coincident and synergistic in time, which is the key stage for triggering early warning.

[0057] In the dynamic window construction module, the microbial incubation cycle parameter refers to the average time required for each type of microorganism to reproduce from contact with a suitable environment to the completion of a generation, used to determine the length of the risk window; the microbial sampling chamber refers to a sealed sampling and analysis space designed specifically for continuous monitoring of microbial concentration in the air, ensuring sample representativeness and detection accuracy; the concentration-increasing sample sequence refers to a group of samples and their data in which the microbial concentration shows a continuous upward trend over multiple consecutive monitoring periods; the rate of increase characteristic refers to the speed of increase of microbial concentration or fluorescence signal within consecutive detection periods, which is a core parameter for determining whether potential risks have occurred; the concentration end change refers to the trend of microbial concentration at the end of the monitoring window compared to the previous sequence, reflecting the microbial growth trend at the end of the risk window; adjusting the sampling frequency refers to automatically increasing or decreasing the sampling interval of the sampling device according to the rate of change of microorganisms within the risk window to adapt to real-time monitoring needs.

[0058] In the early warning linkage output module, the fluctuation status refers to the dynamic performance of microbial concentration, such as increase, decrease, or stagnation, within the current monitoring period, used to determine whether an early warning mechanism is triggered; the linkage trigger standard refers to the criteria for triggering early warning linkage when multiple indicators, such as changes in microbial concentration and environmental parameters, reach or exceed a certain combination of conditions; the audible and visual warning equipment refers to alarm devices installed in the production workshop that can issue risk warnings through sound and light to remind operators to respond in a timely manner; the air circulation equipment refers to facilities used to improve airflow and dilute or purify pollution sources, such as ventilation fans and air purifiers, which are automatically controlled by the system; the microbial node information refers to all microbial monitoring data nodes that are determined to be critical growth, activity changes, or potential risk periods, serving as the basis for source tracing analysis; and the control signal refers to the operation command signal issued based on the monitoring and judgment results, used to link the start and stop of alarms, ventilation, and other related hardware execution units.

[0059] Please see Figure 2 The environmental parameter monitoring module includes:

[0060] The data stream receiving submodule is based on the food production workshop. It analyzes the time series data generated by the temperature and humidity sensors, classifies each sampling point according to its spatial location and collection time, compares the temperature and humidity changes of each sampling point within the same period, and judges the data continuity and the integrity of the sampling point distribution to obtain the periodic monitoring data group.

[0061] Based on the functional division of the production workshop, temperature and humidity sensors were deployed at several key nodes, including the raw material inlet, core processing area, packaging area, and discharge outlet. Each sensor was assigned a unique identification number upon installation, and its corresponding spatial coordinates were determined using the workshop's 3D floor plan. A built-in mapping table was used to bind the number to the location. During data acquisition, each sensor continuously transmitted raw temperature and humidity data every 10 seconds. The data carried a time stamp and sensor number as basic fields. The spatial location was matched against the sensor number, thus grouping the data spatially. For example, sensors with coordinates within 0 to 5 meters on the X-axis were assigned to the inlet area, those within 5 to 20 meters to the processing area, and those above 20 meters to the packaging area. For time-based classification, a monitoring cycle of one hour was used, dividing the daily operating time into multiple fixed cycles. Each cycle formed an independent data segment, and the temperature and humidity data for each sampling point within the cycle were then extracted point-by-point. The system collects temperature variation values ​​and performs difference processing on the maximum and minimum values ​​of all sampling points within the same period to identify the magnitude of the change. Simultaneously, it extracts the average value for horizontal comparison. If the temperature difference between any two points within the same period exceeds 4.0°C, or the humidity difference exceeds 15.0°C, it is marked as having spatial differences and included in the key monitoring area. Next, based on the number of data records for each sampling point within a single period and their chronological order, the system determines whether the data is complete. If a sampling point is missing data for two or more consecutive time points within a period, or if its recording interval is greater than 20 seconds, it is recorded as discontinuous and marked as having missing data within that period. Subsequently, the ratio between the number of sampling points marked as complete and the total number of points is calculated. If it is less than 80.0%, the data for that period is marked as incomplete. Finally, each sampling point within each period is archived separately according to space and time, generating a structured periodic monitoring data set.

[0062] The fluctuation identification submodule is based on the periodic monitoring data set to determine the continuous periodic changes in temperature and humidity at each sampling point, identify segments with changes, compare the characteristics of changes in air disturbance parameters within each segment, mark the time periods of abnormal changes, and obtain the air disturbance feature set.

[0063] First, the temperature and humidity records of each sampling point are extracted sequentially over multiple periods according to the sampling point number. A series of temperature and humidity changes over multiple periods is formed for each sampling point. Then, the temperature difference and humidity difference between adjacent periods are calculated to determine whether the difference meets the fluctuation judgment criteria. If the temperature change amplitude is greater than or equal to 2.0 or the humidity change amplitude is greater than or equal to 8.0, it is marked as a change segment. Further identification of whether the segment has air disturbance characteristics is performed. The identification criteria for disturbance characteristics include: the change segment must exist continuously for more than two periods, the maximum change amplitude must be greater than 4.0 or 10.0 in the whole segment, and there must be at least two temperature and humidity value reversals within the segment, i.e., a jump in the opposite direction, such as an increase followed by a decrease and then an increase. Each change segment that meets the above three conditions is marked, and the start and end time of the disturbance, the change amplitude, the number of fluctuation duration periods, and the number of reversals are recorded. Then, the disturbance markings of all sampling points are cross-analyzed on the same time axis. If more than 30.0% of the sampling points are judged to have disturbances in the same time period, the time period is defined as an air disturbance period and stored in the disturbance feature set.

[0064] The energy trend determination submodule is based on the air disturbance feature set. It simultaneously collects energy absorption data of infrared sensing devices in multiple characteristic bands, determines the changing trend of each moisture channel within the monitoring period, adjusts the sampling interval configuration, and obtains the environmental change parameter set.

[0065] Using the time period recorded by the concentrated air disturbance characteristics as an index, energy absorption data for the corresponding time period is synchronously retrieved from the infrared sensing device. The device records energy values ​​for different bands at a frequency of once per second during operation. Time series data are constructed for each band, and the first and last values ​​of each series are extracted and compared to determine whether the overall change direction of the band is continuous within the disturbance period. The judgment is based on the total change amplitude within the segment. If the energy change amplitude of any band within the time period is greater than 5.0, and the change direction is continuously increasing or continuously decreasing, the band is marked as having a trend. The system identifies the trend characteristics by grouping all bands with the same trend direction into a trend-dominant channel and recording their numbers and variation features. Then, it calculates the difference between the maximum and minimum values ​​of each band during the disturbance period to obtain the overall fluctuation amplitude. If this value exceeds 5.0, the disturbance is considered to have strong fluctuation characteristics. Subsequently, the infrared sensor sampling settings are adjusted, changing the original sampling frequency from once per second to once per two seconds to enhance subsequent monitoring capabilities. The disturbance time period number, trend channel number, energy change direction, total fluctuation amplitude, and adjusted sampling frequency are output as an environmental change parameter set.

[0066] Please see Figure 3 The activity feature recognition module includes:

[0067] The fluorescence signal detection submodule analyzes the fluorescence response signal detected by the air sampling device based on the environmental change parameter set, organizes the signal data sequentially according to the sampling period, compares the speed of signal change in each period, determines the position where the signal curve begins to rise, and obtains the periodic fluorescence response sequence.

[0068] First, raw fluorescence response signal data with timestamps are extracted from the air sampling equipment. Sampling periods are divided into 60-second intervals. Fluorescence signal values ​​within each period are arranged chronologically to construct a time series. This time series is then segmented and processed by period. If a period contains fewer than 6 signal points, it is marked as an abnormal period and removed. This process ensures that each period has a sampling span of at least 60 seconds. After processing, a periodic signal change rate comparison operation is performed. For each period's signal sequence, the signal values ​​corresponding to the start and end times of the period are extracted. The average change rate is obtained by calculating the signal difference divided by the total number of seconds in the period. When comparing the change rates across all periods, 0.2 is used as the threshold value. The rate reference value is used to classify the period. If the average increase in signal value per second exceeds 0.2, it is marked as a fast growth period. If it is less than 0.05, it is marked as a slow growth period. If it is between the two, it is marked as a medium growth period. After the rate classification is completed, the starting position of the signal curve is further determined. The method is to judge its monotonicity by taking three consecutive points as a group within the period. If the signal value of three consecutive points shows an increasing state and the total change exceeds 1.5, the time point corresponding to the first point is marked as the starting point of the rise of the period. If there are multiple points that meet this condition, the earliest time is selected as the starting point of the rise of the period. The data marked as normal, signal processing, rate determination and rise position determination in all periods are integrated to form a periodic fluorescence response sequence.

[0069] The activity change recognition submodule is based on the periodic fluorescence response sequence to determine the change trend of the signal in each period, calculate the signal growth rate within the continuous period, compare the correspondence between the cell activity period parameters and the current signal trend, and obtain the activity period matching sequence.

[0070] First, based on the period number, the average rate of change between adjacent periods in the sequence is extracted, and a growth rate sequence of adjacent periods is constructed. Then, the difference between the rate of each period and the rate of the previous period is compared. If the current period's rate is greater than the previous period's rate by more than 0.1, it is considered a state of accelerated growth; if it is less than the previous period's rate by less than 0.1, it is considered a state of slowed growth; if the difference is between ±0.1, it is considered a state of leveled growth. This comparison is performed for each period sequence, and a trend label is assigned to each period. After trend labeling, a cell activity cycle parameter is introduced to determine the corresponding trend. This parameter is set according to the microbial species type, for example... For example, if a certain bacterial species has a reproductive cycle of 3 hours under temperature and humidity conditions of 20℃ to 30℃, and the active growth period is set to the first 1 hour to the second hour, and the cycle number is converted into the corresponding sampling time with the hour as the cycle unit, it is determined whether the current signal growth trend falls within the active growth period of the bacterial species. If the sampling time of a certain cycle overlaps with the active cycle of the bacterial cell, and the signal growth trend is marked as accelerated growth or continuous growth, it is determined to be an active trend matching cycle. If the signal growth trend is slowing down or flat, or does not fall within the active period, it is not matched. In this way, all cycle data are compared, all matching items are output, and an active cycle matching sequence is generated.

[0071] The growth node extraction submodule is based on the active cycle matching sequence to screen for cycles with continuously increasing signals, determine the continuity of microbial concentration changes within the cycle, classify and label time points with change characteristics, and obtain growth characteristic node groups.

[0072] Periods that are identified as having a matching activity trend in two or more consecutive periods are selected. The selection criteria are a continuity greater than 2 and an average signal growth rate greater than 0.15. After selection, microbial concentration change characteristic data within each matching period are further extracted. The concentration change data is obtained by converting the fluorescence response signal intensity. The conversion rule uses the difference in fluorescence value per unit time as the concentration change index. If the difference in concentration change between the first half and the second half of the period exceeds 2.0, the period is marked as having change characteristics. Time point classification processing is performed on the marked periods. The method is to divide the signal intensity into multiple time periods in 30-second intervals within the period and judge the change rate of each interval. If the signal growth rate in a single interval exceeds 0.25 and the growth trend is consistent between two adjacent time periods, the end time point of the previous interval is marked as the characteristic time point. If multiple intervals meet this condition, they are merged into continuous node intervals. All time points with change characteristics are archived according to the period number to form growth characteristic node groups.

[0073] Please see Figure 4 The triggering event analysis module includes:

[0074] The node synchronization discrimination submodule is based on the growing feature node group. It determines the synchronization status of the time position of each feature node with the temperature and humidity changes within the corresponding sampling period, compares the temporal correlation between each node and the changes in environmental parameters, identifies nodes with highly consistent trends and combinations of environmental data, and obtains a node-environment temporal matching set.

[0075] Read the timestamp information recorded in each feature node and extract its corresponding sampling period number. Then, retrieve the archived temperature and humidity data sequence within that period. Align the node occurrence time with the time points in the temperature and humidity change sequence. Expand the window forward and backward by 60 seconds with a step size of 10 seconds. Extract 6 sets of temperature and humidity values ​​before and after the node. Determine whether there is a trend change in temperature and humidity within the sampling period before and after the node occurrence. The temperature trend is judged by whether there are three consecutive time periods of increase or decrease. If three consecutive changes occur in the same direction and the total change amplitude exceeds 2.0, a temperature trend change is determined. The humidity is judged in the same way, with a threshold set to a change amplitude exceeding 6.0. If both temperature and humidity conditions are met simultaneously, it is considered a synchronous change. For each periodic node, if only one condition is met, it is considered partially synchronized; if none are met, it is considered asynchronous. The synchronization status of each node is marked as "synchronized," "partially synchronized," or "asynchronous." After aligning the periodic data within a single node, the temperature and humidity change direction of each node is compared horizontally with the trend direction of other nodes within the same time period. If the occurrence times of multiple nodes all fall within a certain continuous time interval, and their respective temperature and humidity change directions are consistent, then the nodes and environmental parameters are recorded as having trend consistency and are grouped together. The node numbers with consistent trends are paired with their corresponding temperature and humidity change directions, and their occurrence time intervals are recorded as indexes. All node data determined to be trend consistent are output in a structured manner to obtain the node-environment time-series matching set.

[0076] The log period extraction submodule is based on the node environment time sequence matching set to determine the humidifier and air conditioner parameter change behavior in the workshop adjustment log within the associated time period, compare the operation activity of the equipment adjustment record with the node matching time period, identify the log segment with frequent operation, and obtain the parameter disturbance segment set.

[0077] The system reads the start and end times of the time segment marked in each record of the matching set, compares the time segment with the workshop environmental control equipment log data with minute-level precision. The log data focuses on humidifiers and air conditioners; each parameter adjustment operation records the operation time, equipment number, parameter type (e.g., temperature setpoint, humidity setpoint), operation value, and operator information. All log records are sorted in ascending order of time, and all log data overlapping with the time range of each trend segment are extracted. If a log time falls within the start and end time of the trend segment, it is recorded as a matching log record. Subsequently, the frequency of log operations within each trend segment is statistically analyzed. If a humidifier or air conditioner is operated more than 3 times, it is marked as an active operation segment. If the number of operations is between 1 and 3 times, it is marked as moderately active. If there is no operation, it is marked as a silent segment. At the same time, the change range of the device's adjustment parameters is calculated. If a device experiences a change in set value greater than 2.0 more than 2 times within a trend segment, it is also judged as a frequent disturbance as an auxiliary criterion. In addition, the direction of parameter adjustment is checked. If two consecutive adjustments are made in the upward direction within a certain period of time, it is marked as "continuously increasing disturbance". If one adjustment is increased and the other is decreased, it is marked as "oscillating disturbance". All trend segments that meet any of the above disturbance characteristics will be classified into the parameter disturbance segment set.

[0078] The response segment positioning submodule, based on the parameter disturbance segment set, calculates the time synchronization length, temperature and humidity parameter fluctuation amplitude, and number of feature nodes between each segment using the following formula:

[0079] ;

[0080] Obtain the linkage strength factor of the response section By filtering out sections with high linkage intensity, the linkage response interval is obtained, among which... Indicates the total number of overlapping segments. Indicates the first Time synchronization length of each segment Indicates the first The fluctuation range of temperature and humidity parameters in each section Indicates the first The number of matching feature nodes in each segment;

[0081] The response segment linkage strength factor is a numerical indicator that reflects the degree of temporal synchronization and spatial coupling between the microbial growth characteristic nodes and temperature and humidity parameter fluctuations within each response segment. This factor comprehensively considers the data performance of the number of characteristic nodes, the length of time synchronization, and the amplitude of temperature and humidity changes within the segment. It is used to quantitatively determine the strength of the synergistic relationship between environmental changes and microbial activities in different monitoring segments. The larger the factor value, the stronger the linkage between changes in microbial activity and temperature and humidity disturbances within the segment.

[0082] Identify all active parameter segments and, based on the node environment time-series matching set, calculate three key parameters for each segment: time synchronization length. Temperature and humidity parameter fluctuation range and the number of matching feature nodes The time synchronization length This indicates the overlap between the occurrence time of the feature node and the time period of temperature and humidity abrupt changes, counted in minutes. For example, the original durations corresponding to segments D1 to D4 are 12, 15, 10, and 8 minutes respectively; the fluctuation range of temperature and humidity parameters. This indicates the average degree of change in environmental parameters within each section, which can be calculated as the average of the sum of the maximum and minimum temperature and humidity differences in each section. The original data are 1.8, 2.1, 1.2, and 1.5 (unit: ℃ / %RH); Number of matching feature nodes. The number of active microbial nodes that exhibit a trend consistent with environmental fluctuations within this timeframe is 3, 5, 2, and 1, respectively; these are then uniformly mapped using minimum-maximum normalization. The interval, after normalization, is as follows:

[0083] ;

[0084] ;

[0085] ;

[0086] Then, substitute the above results into the calculation formula, where That is, the number of overlapping segments, which is calculated as follows:

[0087] The molecular part is:

[0088] ;

[0089] ;

[0090] The denominator is:

[0091] ;

[0092] The linkage strength factor is obtained as follows:

[0093] ;

[0094] The linkage intensity is divided into the following three sections:

[0095] when When defined as a weakly linked segment, it indicates that there is a lack of significant temporal or quantitative synchronization between microbial characteristic nodes and environmental fluctuation events;

[0096] when At this time, it is defined as a medium linkage segment, indicating that there are some parameter coupling characteristics but a stable temporal correlation structure has not yet been formed;

[0097] when When the time is defined as a strong linkage segment, it means that the time synchronization length is sufficient, the parameter fluctuation amplitude is significant, and the matching nodes are dense. This situation corresponds to the key monitoring objects.

[0098] Current value The fact that it is clearly located in a highly correlated zone indicates that there is a high degree of synergy between environmental disturbances and microbial growth events in this zone. By using threshold division logic, the current zone is identified and marked as a linked response interval, providing a basis for subsequent sampling frequency adjustment and early warning linkage output.

[0099] Please see Figure 5 The dynamic window building module includes:

[0100] The incubation cycle extraction submodule analyzes the temperature and humidity sequences based on the linkage response interval, compares the combined temperature and humidity changes of each monitoring segment, determines whether the trend of the monitoring data changes conforms to the continuous characteristics of the incubation period of the strain, identifies the corresponding data segments, and obtains the microbial incubation cycle parameters.

[0101] Extract the start and end times recorded in each linkage response interval. Extract the temperature and humidity data sequences within this time range in minutes, constructing two time series vectors of equal length. Then, pair the temperature and humidity data at each time point to form a joint data pair. Next, perform trend judgment on the data pairs. The judgment criterion is whether temperature and humidity simultaneously increase or decrease within three consecutive time points. If the temperatures at three consecutive time points are 28.1, 28.6, and 29.0, and the humidity is 62.0, 64.5, and 67.0, then this segment is considered to have a consistent upward trend. Using five-minute intervals as monitoring segments, calculate the matching of the direction and magnitude of temperature and humidity changes within each segment. If the total temperature increase is greater than 1.5 and the total humidity increase is greater than 5 within the monitoring segment... If a value of .0 indicates a joint trend, the segment is considered to have a certain degree of change. If at least two consecutive monitoring segments meet this condition, the time period is preliminarily identified as a potential incubation segment. Subsequently, the segment is compared with the known incubation temperature and humidity ranges for bacterial species. For example, the incubation of a certain bacterium requires a temperature between 27.0 and 32.0°C and a humidity between 60.0 and 75.0%. The mean temperature and mean humidity in each monitoring segment are judged according to the range. If both are within the range and the trend is continuous, the monitoring segment is judged to meet the characteristics of continuous incubation. All monitoring segments that meet the above temperature and humidity range judgment and trend continuity will be integrated and merged to form a complete incubation cycle segment. Data such as start time, end time, duration, average temperature, average humidity, and trend type are extracted from the segment and archived, and output as microbial incubation cycle parameters.

[0102] The incremental sequence identification submodule compares the fluorescence intensity sequences of each monitoring period based on the microbial incubation cycle parameters, calculates the fluorescence change amplitude within the continuous period, identifies the data segment with continuously increasing change amplitude, judges the sample sequence of the continuously increasing segment, and obtains the concentration-increasing sample sequence.

[0103] Obtain the raw fluorescence intensity data within the marked start and end time segments of the corresponding incubation cycle parameters, and divide it into several fluorescence intensity cycle segments, one per hour. Then, extract the starting and ending fluorescence values ​​of each cycle segment, and calculate the total change within each cycle segment, i.e., the result of subtracting the starting value from the ending value. If the change value is positive and greater than 1.0, it is considered an effective growth cycle. Then, perform a continuity judgment on all effective growth cycles. If three consecutive cycles meet the above growth conditions, and the fluorescence change value of each cycle shows an increasing trend compared to the previous cycle, i.e., the change value of the first cycle is 1.2, the second is 1.7, and the third is 2.3, then it is judged as an increasing trend. The process involves marking each segment with its cycle number, start time, cycle variation, and fluorescence value difference. It then determines whether the fluorescence data covered by the increasing segment is continuous and without gaps. If a cycle has missing data or fewer than 6 samples, that cycle is discarded. If all increasing cycles are fully sampled and the average growth rate exceeds 1.5%, the segment is included in the concentration increasing determination range. A representative sample sequence is extracted from this sequence, containing the fluorescence change curve for each cycle, sampling timestamp, variation range, growth rate, continuity, and overlap with incubation parameters. All qualifying sample cycles are structured and output to form a concentration increasing sample sequence.

[0104] The trend feature summarization submodule, based on the concentration-increasing sample sequence, calculates the concentration shift between the endpoint of each segment and the previous period, optimizes the joint processing of concentration change and fluorescence rate, and uses the following formula:

[0105] ;

[0106] Get the dynamic growth window identifier value To determine the strength of the trend feature induction, the sampling interval is adjusted to obtain a dynamically growing window group, where... Indicates the number of consecutively increasing periods. Indicates the first The change in concentration in segment is the first segment of the sample sequence. The concentration difference between the end and the beginning of each increasing cycle. Indicates the first The dimensionless fluorescence rate ratio of the segments is calculated as follows: , Indicates the first Segment fluorescence change rate, This represents the average fluorescence rate of change across all segments. Indicates the first The normalized concentration value of the period is the number of periods in the sample sequence. The concentration of microorganisms monitored in each cycle, Indicates the first The normalized concentration value for the period is the first... The concentration of microorganisms monitored in each cycle;

[0107] The dynamic growth window identifier is a composite index used to quantify the increasing trend and intensity of microbial concentration within a continuous monitoring period. By mathematically quantifying the growth rate, change amount, and trend consistency of microbial concentration within each continuous increasing segment, the increasing trend of multiple periods is summarized into a measurable value, reflecting whether the microbial concentration within this time window is in a state of continuous growth and strong change characteristics.

[0108] Set the number of sample segments to Trend characteristics of microbial concentration sequences over continuous monitoring periods are identified, and the concentration difference between the endpoint of each period and the previous period is calculated to form a concentration change sequence. And extract the fluorescence change rate of the corresponding segment. Subsequently, respectively , and concentration sequence The range was normalized to obtain the normalized concentration change. Normalized concentration sequence and the dimensionless fluorescence rate ratio The joint index of trend characteristics was calculated using a formula, where the first term measures the coupling trend between normalized concentration change and fluorescence response, and the second term reflects the fluctuation intensity of the normalized concentration sequence. Calculation results. It is a dimensionless composite index that can be used to determine dynamic trends.

[0109] Taking an actual monitoring sequence as an example, let the original concentration sequence be... The change in concentration is The fluorescence rate is The concentration sequence was normalized by range, with a maximum value of 78, a minimum value of 52, and a range of 26, resulting in the normalized concentration sequence. Normalizing the concentration change, the maximum value is 9, the minimum value is 8, and the range is 1, yielding the following results. Normalizing the fluorescence rate, the maximum value is 2.1, the minimum value is 1.6, and the range is 0.5. .

[0110] Substituting into the formula, the first term is:

[0111] ;

[0112] The second item is:

[0113] ;

[0114] The calculation shows that:

[0115] ;

[0116] Based on trend judgment criteria:

[0117] when It was determined to be "insignificant growth".

[0118] when At that time, the trend was "awaiting confirmation".

[0119] when This is then considered an "effective growth window".

[0120] because The sample sequence falls within the range above the threshold, therefore it can be identified as an effective growth window, indicating that the current microbial concentration is showing a continuous and significant upward trend.

[0121] Please see Figure 6 The early warning linkage output module includes:

[0122] The concentration fluctuation judgment submodule is based on a dynamic growth window group. It analyzes the changes in microbial concentration in each monitoring period within the dynamic growth window group, judges whether the concentration rises, falls, or remains flat in each period, compares the trend of concentration changes in each stage in continuous periods, classifies the fluctuation status, and obtains the microbial concentration fluctuation status.

[0123] Extract all monitoring cycle numbers and corresponding microbial concentration values ​​recorded in the window group, and construct a complete concentration time series dataset in chronological order. Then, divide each monitoring cycle into independent analysis units, and assess the concentration changes within each cycle. The assessment method involves extracting the microbial concentration values ​​at the start and end of the cycle and calculating their difference. If the difference is greater than 0.5, it is considered an upward trend; if the difference is less than -0.5, it is considered a downward trend; and if the difference is between -0.5 and 0.5, it is considered a stable trend. After assessing the state of each cycle, compare the trends of two or more consecutive cycles. If both cycles show an increase, it is classified as continuous growth; if one increases and the other decreases, it is considered an oscillating change. If both cycles remain flat, it is classified as a stable state. Each group of three-cycle segments is classified, and adjacent cycle states are labeled with trend tags. The tag types are divided into four categories: "continuous rise", "continuous fall", "alternating fluctuations", and "continuous flatness". The total number of occurrences and proportions of each trend type in the entire dynamic window group are further counted, and the trend type with the highest frequency is marked as the dominant fluctuation trend of the window group. For example, if a window group contains 6 cycles, 4 cycles are judged to be rising and 2 are flat, then the dominant trend of the group is "continuous rise". The concentration state of each cycle, the trend category of each pair of consecutive cycles, and the dominant trend are classified and recorded in a structured manner, and the microbial concentration fluctuation state is output.

[0124] The linkage condition matching submodule compares the correlation between the microbial concentration fluctuation state and the linkage triggering standard within the monitoring period to determine whether the parameter combination within the period meets the triggering requirements and obtains linkage judgment matching information.

[0125] First, the temperature and humidity data for each monitoring cycle are loaded along with the concentration fluctuation status of the current cycle, and mapped one-to-one according to the cycle number. Then, the pre-set linkage triggering criteria are read. This criterion includes three parameter combinations: a concentration change greater than 1.5, a temperature between 28.0 and 32.0, and a humidity between 65.0 and 80.0. Meeting any two of these criteria is sufficient to trigger the state. Each cycle is compared one by one. First, it is determined whether the concentration change is greater than 1.5. If the starting concentration of the cycle is 3.1 and the ending concentration is 5.0, the difference is 1.9, satisfying the first condition. Then, the temperature and humidity within that cycle are determined. If the average value is within the set range, and the temperature is 29.2°C and the humidity is 72.5°C, both values ​​are within the corresponding range, then the cycle is recorded as meeting all three linkage conditions. If a cycle only meets the concentration and humidity conditions but the temperature is not within the range, the cycle is also considered to meet the linkage triggering conditions. The number of conditions met in each cycle is counted and marked as whether the linkage standard is met. If the number of conditions met in a cycle is greater than or equal to 2, then the cycle is written into the linkage matching success sequence. The matching information records fields such as concentration fluctuation type, average temperature and humidity, judgment result and number of conditions met. All cycles that are matched successfully are sorted by time and output to form linkage judgment matching information.

[0126] The signal archiving and output submodule analyzes temperature and humidity data, microbial concentration change nodes and equipment control commands based on linkage judgment and matching information, determines the start-up sequence of the audible and visual warning equipment and the air circulation equipment, archives the current cycle monitoring and control data information, and obtains the early warning linkage trigger signal.

[0127] The system retrieves temperature and humidity records, microbial concentration change nodes, and whether any equipment control commands have been issued within the start and end time periods of each matching cycle. Information is organized by cycle number, and environmental parameters and concentration node data are integrated according to time sequence. Then, the control logs of the audible and visual alarm devices and air circulation devices are queried to determine if there are any control signal trigger records within each matching cycle. If a control signal is issued within 60 seconds of the cycle start, it is marked as "synchronous start"; if the control signal appears within 60 seconds of the cycle end, it is marked as "delayed start"; if no control signal is issued within the cycle, it is marked as "no linkage." This is based on the device number and signal type. The alarm commands of the audible and visual devices and the start / stop commands of the air equipment are recorded by comparing with the timestamps. The response times of each device are extracted for the start-up timing judgment. If the audible and visual device responds more than 30 seconds before the air equipment, it is considered a "pre-response", otherwise it is a "post-response". If the interval between the two is no more than 30 seconds, it is a "synchronous response". The above information is organized into a set of data items for each cycle, recording the cycle number, average temperature and humidity, concentration change points, status of the audible and visual devices, status of the air equipment, control timing labels and signal triggering methods. The information is uniformly archived in the monitoring and control data warehouse and structured as early warning linkage trigger signals for automated recording and reporting.

[0128] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. An intelligent monitoring and early warning system for the food production environment, characterized in that, The system includes: The environmental parameter monitoring module is based in the food production workshop. It analyzes real-time temperature and humidity data streams, screens temperature and humidity fluctuations caused by air flow, judges changes in infrared spectral moisture energy, dynamically adjusts the sampling cycle, and obtains a set of environmental change parameters. The activity feature identification module analyzes fluorescence response signals based on the environmental change parameter set, determines the rate of change of microbial reaction curves, compares the consistency of activity cycles and trends, identifies concentration growth samples, and obtains growth feature node groups. Based on the growth feature node group, the triggering event analysis module determines the synchronization of the feature node with the current temperature and humidity on the time axis, identifies the parameter change cycle in the adjustment log, locates the synchronous segment of the damp heat response, and obtains the linkage response interval. The dynamic window construction module analyzes the corresponding bacterial incubation cycle based on the linkage response interval, identifies the increasing samples in the continuous monitoring cycle, judges the rise rate characteristics, adjusts the sampling frequency, and obtains a dynamic growth window group. The early warning linkage output module, based on the dynamically growing window group, judges the microbial concentration fluctuation in the current monitoring period, compares the detection results with the linkage triggering standard, and obtains the early warning linkage triggering signal.

2. The intelligent monitoring and early warning system for the food production environment according to claim 1, characterized in that, The environmental change parameter set includes air state characteristics, sensor monitoring area distribution, and abnormal fluctuation records. The growth characteristic node group includes the bacterial community growth starting point, concentration change trend point, and activity peak identifier. The linkage response interval includes linkage discrimination index, synchronous response period, and control boundary attribute. The dynamic growth window group includes growth window number, incubation characteristic differentiation, and data time series label. The early warning linkage trigger signal includes risk response identifier, linkage execution parameter, and record archiving identifier.

3. The intelligent monitoring and early warning system for the food production environment according to claim 1, characterized in that, The environmental parameter monitoring module includes: The data stream receiving submodule is based on the food production workshop. It analyzes the time series data generated by the temperature and humidity sensors, classifies each sampling point according to its spatial location and collection time, compares the temperature and humidity changes of each sampling point within the same period, and judges the data continuity and the integrity of the sampling point distribution to obtain the periodic monitoring data group. The fluctuation identification submodule, based on the periodic monitoring data set, determines the continuous periodic temperature and humidity changes at each sampling point, identifies segments with changes, compares the characteristics of air disturbance parameter changes within each segment, marks the time periods of abnormal changes, and obtains an air disturbance feature set. The energy trend determination submodule, based on the air disturbance feature set, simultaneously collects energy absorption data of multiple feature bands, determines the changing trend of each moisture channel within the monitoring period, adjusts the sampling interval configuration, and obtains the environmental change parameter set.

4. The intelligent monitoring and early warning system for the food production environment according to claim 1, characterized in that, The activity feature recognition module includes: The fluorescence signal detection submodule analyzes the detected fluorescence response signal based on the set of environmental change parameters, organizes the signal data sequentially according to the sampling period, compares the rate of signal change in each period, determines the position where the signal curve begins to rise, and obtains the periodic fluorescence response sequence. The activity change recognition submodule determines the change trend of each periodic signal based on the periodic fluorescence response sequence, calculates the signal growth rate within consecutive periods, compares the correspondence between the cell activity period parameters and the current signal trend, and obtains the activity period matching sequence. The growth node extraction submodule, based on the active cycle matching sequence, filters out cycles with continuously increasing signals, determines the continuity of microbial concentration changes within the cycle, classifies and marks time points with change characteristics, and obtains growth characteristic node groups.

5. The intelligent monitoring and early warning system for the food production environment according to claim 1, characterized in that, The triggering event analysis module includes: The node synchronization discrimination submodule, based on the growth feature node group, determines the synchronization status of the time position of each feature node with the temperature and humidity changes within the corresponding sampling period, compares the temporal correlation between each node and the changes in environmental parameters, identifies the combination of nodes and environmental data with a trend correlation coefficient greater than a preset threshold, and obtains the node-environment temporal matching set. The log period extraction submodule, based on the node environment time series matching set, determines the environmental adjustment equipment parameter change behavior in the workshop adjustment log within the associated time period, compares the equipment adjustment record with the operation activity of the node matching time period, identifies the log segment with operation change, and obtains the parameter disturbance segment set. The response segment positioning submodule calculates the time synchronization length, temperature and humidity parameter fluctuation amplitude, and number of feature nodes between each segment based on the parameter disturbance segment set, using the following formula: ； Obtain the linkage strength factor of the response section By filtering out sections with high linkage intensity, the linkage response interval is obtained, among which... Indicates the total number of overlapping segments. Indicates the first Time synchronization length of each segment Indicates the first The normalized index of the temperature and humidity parameter fluctuation range of each section Indicates the first The number of matching feature nodes in each segment.

6. The intelligent monitoring and early warning system for the food production environment according to claim 1, characterized in that, The dynamic window construction module includes: The incubation cycle extraction submodule analyzes the temperature and humidity sequences based on the linkage response interval, compares the combined temperature and humidity changes of each monitoring segment, determines whether the trend of the monitoring data changes conforms to the continuous characteristics of the incubation period of the strain, identifies the corresponding data segment, and obtains the microbial incubation cycle parameters. The incremental sequence identification submodule compares the fluorescence intensity sequences of each monitoring period based on the microbial incubation cycle parameters, calculates the fluorescence change amplitude within the continuous period, identifies the data segment with continuously increasing change amplitude, determines the sample sequence of the continuously increasing segment, and obtains the concentration-increasing sample sequence. The trend feature summarization submodule calculates the concentration shift between the endpoint of each segment and the previous period based on the concentration-increasing sample sequence, optimizes the joint processing of concentration change and fluorescence rate, obtains the dynamic growth window identifier value, judges the strength of trend feature summarization, adjusts the sampling interval, and obtains the dynamic growth window group.

7. The intelligent monitoring and early warning system for the food production environment according to claim 1, characterized in that, The early warning linkage output module includes: The concentration fluctuation judgment submodule analyzes the changes in microbial concentration in each monitoring period within the dynamic growth window group based on the dynamic growth window group, judges the rise, fall or stagnation of concentration in each period, compares the trend of concentration changes in each stage in the continuous period, classifies the fluctuation state, and obtains the microbial concentration fluctuation state. The linkage condition matching submodule compares the microbial concentration fluctuation state with the corresponding relationship of the linkage triggering standard within the monitoring period to determine whether the parameter combination within the period meets the triggering requirements and obtains linkage judgment matching information. Based on the linkage determination and matching information, the signal archiving and output submodule analyzes temperature and humidity data, microbial concentration change nodes and equipment control commands, determines the start-up sequence of the audible and visual warning device and the air circulation device, archives the current cycle monitoring and control data information, and obtains the early warning linkage trigger signal.

8. The intelligent monitoring and early warning system for the food production environment according to claim 1, characterized in that, The data stream refers to the continuous time series of raw environmental data such as temperature and humidity continuously collected and transmitted by the temperature and humidity sensor, and the fluorescence response signal refers to the fluorescence intensity signal generated by microorganisms or other biological components under an excitation light source.

Images