Forestry seedling raising monitoring method and system based on Internet of Things

Abstract

The invention relates to the technical field of forestry monitoring, in particular to a forestry seedling raising monitoring method and system based on the Internet of Things, and the method comprises the following steps: obtaining soil humidity, temperature and illumination data of a monitoring point, positioning an abnormal fluctuation range, generating a fluctuation boundary diagram, recognizing an illumination and humidity collaborative change vector, and counting the frequency; the collaborative distribution information is extracted to generate a synchronism index, synchronism deviation is analyzed, a deviation distribution diagram is generated, the sampling density is subjected to clustering analysis, and a dynamic change data set is generated through normalization. According to the method, a fluctuation extreme point is positioned through adjacent time sequence difference operation, abnormal fluctuation characteristics of seedling environment parameters are captured, the real-time performance and accuracy of data anomaly detection are improved, illumination and humidity collaborative change direction vectors are identified, collaborative frequencies are subjected to aggregation statistics according to segments, and dynamic relevance of environment factors is quantified. The coupling relation between the monitoring points is accurately identified, the monitoring points with synchronism deviating from the threshold value are screened, the space-time heterogeneity characteristics are revealed, and the monitoring point layout is optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of forestry monitoring, and particularly to a forestry seedling cultivation monitoring method and system based on the Internet of Things. Background Technique

[0002] The technical field of forestry monitoring includes the continuous observation, recording, and analysis of forest resources, growth environment, biodiversity, and forestry activities. The core content of this technical field is to collect and analyze data on key indicators such as forest soil, water, climate, growth status, and pests and diseases through various sensing means and observation methods to support forestry resource management, ecological protection, and forest cultivation planning. The overall technical system of forestry monitoring covers remote sensing image analysis, field investigation methods, environmental factor monitoring techniques, and information integration and analysis means, and is widely used in various aspects such as forest resource surveys, forest disaster early warnings, tree growth assessments, and ecological quality evaluations, forming a forestry scientific management system based on data-driven.

[0003] Among them, the forestry seedling cultivation monitoring method refers to a technical solution for observing and recording key growth parameters and environmental elements in the whole process of seedling cultivation. It mainly regularly measures parameters such as soil humidity, temperature, light intensity, and seedling height of seedlings in the nursery, and analyzes the change of seedling state by comparing with the set growth cycle standard. The specific methods used include using temperature and humidity sensors to obtain soil information, using light intensity meters to record light changes, using scale measuring tools to record the change of seedling height at fixed points, and at the same time combining historical records to organize time series data and determine the growth stage. This monitoring method uses field observation combined with basic measuring tools as means to achieve continuous recording and stage diagnosis of the forestry seedling cultivation process.

[0004] The prior art relies on regular measurements and basic tools to record data, resulting in the monitoring results being limited to static parameters at discrete time points and making it difficult to capture the continuous fluctuation characteristics of environmental factors. For example, temperature and humidity sensors and light intensity meters only record instantaneous values and cannot identify the abnormal fluctuation ranges of adjacent time series, missing the impact of short-term drastic changes on seedling growth. When integrating time series data in historical records tables, co-variation analysis is lacking, resulting in the dynamic correlation of environmental factors such as light and humidity not being quantified and making it difficult to evaluate the impact of multi-factor coupling on seedling cultivation. The existing methods do not introduce response time distribution and coupling strength indicators and cannot quantify the synchronous difference in the changes of environmental factors at different monitoring points. For example, the problem of lag in humidity and light responses due to micro-environment differences in different areas within the same nursery is not identified. Field surveys and scale measurement tools rely on manual operations, with a fixed sampling point density that cannot be dynamically adjusted, ignoring the monitoring blind spots caused by environmental heterogeneity in local areas. For example, traditional methods are prone to missing small-scale abnormal fluctuations in sparse monitoring point areas, while redundant data is generated in high-density areas. The prior art does not construct a dynamic change data set, resulting in the monitoring data being difficult to support the accurate diagnosis of the spatio-temporal heterogeneity of the seedling cultivation environment and affecting the timeliness of disaster warning and resource scheduling. Summary of the Invention

[0005] An object of the present invention is to solve the deficiencies existing in the prior art and to propose a forestry seedling cultivation monitoring method and system based on the Internet of Things.

[0006] To achieve the above object, the present invention adopts the following technical solutions: A forestry seedling cultivation monitoring method based on the Internet of Things, comprising the following steps:

[0007] S1: Obtain the soil humidity, temperature, and light intensity data of the monitoring points in the seedling cultivation environment, perform adjacent time series difference operations, locate the fluctuation extreme points and determine the starting position of the downward trend, determine the abnormal fluctuation range, and generate a discrete distribution map of the fluctuation boundary.

[0008] S2: Based on the discrete distribution map of the fluctuation boundary, read the abnormal fluctuation sequence, identify the co-variation direction vector of light and humidity, perform co-variation counting on the vector and aggregate it by segment, and count the co-variation frequency of light and humidity to obtain the co-variation frequency data of abnormal fluctuations.

[0009] S3: According to the co-variation frequency data of abnormal fluctuations, extract the co-variation distribution information of the monitoring points, combine the response time distribution of the environmental factors at the monitoring points, identify the coupling strength index of the environmental factors, and generate the synchronization index of the environmental factors at the monitoring points.

[0010] S4: Based on the synchronization index of the environmental factors at the monitoring points, screen the set of monitoring points whose synchronization deviates from the median threshold, analyze the corresponding differences between the sampling point interval and the response time interval, and generate a synchronization deviation distribution map of the environmental factors.

[0011] As a further solution of the present invention, the discrete distribution map of the fluctuation boundary includes the distribution of the positions of the fluctuation extreme points, the distribution of the starting points of the downward trend, and the abnormal fluctuation range. The abnormal fluctuation co-frequency data includes the co-variation frequency of the fluctuation segments, the number of co-variations of the vectors, and the characteristics of the change in the fluctuation direction. The synchronization index of the environmental factors at the monitoring points includes the co-variation distribution information, the response time distribution, and the coupling intensity index of the environmental factors. The distribution map of the synchronization deviation of the environmental factors includes the time distribution of the sampling points, the average value of the sampling point intervals, and the corresponding difference distribution of the monitoring points.

[0012] As a further solution of the present invention, the specific steps for obtaining the discrete distribution map of the fluctuation boundary are as follows:

[0013] S111: Obtain the soil humidity, temperature, and light intensity data of the monitoring points in the seedling-raising environment, perform adjacent time series difference operations on the continuous data, identify the points where the difference changes sign, and extract the points where the two points before and after the current data point have reverse differences as the fluctuation extreme points to obtain the position sequence of the extreme points.

[0014] S112: According to the position sequence of the extreme points, for each extreme point and the adjacent data sequence, calculate the continuous difference and determine whether it is less than zero, mark the starting point of the downward trend, and combine the extreme points to form a paired structure to obtain the sequence of the abnormal fluctuation range.

[0015] S113: Based on the sequence of the abnormal fluctuation range, group by the starting time of the fluctuation segments, calculate the start and end differences, the decline interval, and the average decline rate of each group of fluctuation segments, integrate the boundary characteristics of the fluctuation segments, and generate the discrete distribution map of the fluctuation boundary.

[0016] As a further solution of the present invention, the specific steps for obtaining the abnormal fluctuation co-frequency data are as follows:

[0017] S211: Based on the discrete distribution map of the fluctuation boundary, read the abnormal fluctuation subsequences within the corresponding time range, identify the index sequence of each fluctuation in ascending order of time, and perform reverse sorting on the light and humidity data sequences to obtain the light and humidity difference vector sequences.

[0018] S212: According to the light and humidity difference vector sequences, extract the sign changes of adjacent elements in the differences. If it is less than zero, it is regarded as a co-variation event, and count the number of co-variations in each index segment to obtain the co-variation count sequence.

[0019] S213: According to the co-variation count sequence, combined with the time span information of each fluctuation segment, calculate the co-variation frequency per unit time within each fluctuation segment, count the corresponding frequencies of the fluctuation segments, and establish the abnormal fluctuation co-frequency data.

[0020] As a further solution of the present invention, the steps for obtaining the environmental factor synchronization index of the monitoring points are specifically as follows:

[0021] S311: According to the abnormal fluctuation collaborative frequency data, map and aggregate by monitoring point number, extract the associated abnormal fluctuation numbers and collaborative frequencies, calculate the collaborative change distribution of each monitoring point in the abnormal fluctuations in groups, and generate a monitoring point collaborative change distribution sequence;

[0022] S312: Based on the monitoring point collaborative change distribution sequence, extract the monitoring point response time sequence, calculate the fluctuations corresponding to the distribution information at the time points, evaluate the time distribution fluctuations of each monitoring point response sequence, and generate a monitoring point response time distribution sequence;

[0023] S313: According to the monitoring point response time distribution sequence, combine the short-term environmental factor distribution of the monitoring points, statistically calculate the mean value, energy concentration degree and standard deviation, evaluate the environmental factor coupling structure, and calculate the coupling intensity index to obtain the environmental factor synchronization index of the monitoring points.

[0024] As a further solution of the present invention, the steps for obtaining the environmental factor synchronization deviation distribution map are specifically as follows:

[0025] S411: Based on the environmental factor synchronization index of the monitoring points, identify the synchronization value set by monitoring point number, extract the quartiles and set the abnormal deviation threshold, screen the monitoring point index values, and output the set of abnormal monitoring point numbers;

[0026] S412: According to the set of abnormal monitoring point numbers, extract the difference between consecutive sampling point intervals of the monitoring points and standardize it, calculate the ratio of the corresponding time of the fluctuation segment to the sampling misalignment, aggregate and calculate the average value of the misalignment ratio sequence of each monitoring point to generate a monitoring point alignment deviation mean sequence;

[0027] S413: Based on the monitoring point alignment deviation mean sequence, sort the average alignment deviations by monitoring point number, plot the synchronization deviation trajectory map with the time difference as the ordinate and the monitoring point number as the abscissa, calculate the corresponding deviation between the sampling time and interval of the monitoring points, and use the deviation as the discrete trajectory points of the monitoring points to generate the environmental factor synchronization deviation distribution map.

[0028] As a further solution of the present invention, the method further includes step S5:

[0029] S5: According to the environmental factor synchronization deviation distribution map, perform clustering analysis on the density distribution of the reorganized sampling points, statistically calculate the sampling density gradient of the clustering region, perform difference operation with the reference density and perform normalization processing to generate a dynamic change data set of the seedling raising environment;

[0030] The dynamic change dataset of the seedling raising environment includes the time window clustering area, the change rate of sampling density gradient, and the normalized difference area distribution record.

[0031] As a further solution of the present invention, the steps for obtaining the dynamic change dataset of the seedling raising environment are specifically as follows:

[0032] S511: According to the environmental factor synchronization deviation distribution map, reorder all sampling points on the time axis, calculate the change trend of sampling point density, and count the change rate of sampling points within a unit time to obtain the sliding window sampling density gradient;

[0033] S512: Based on the sliding window sampling density gradient, extract the sampling density within the time window and perform a difference operation with the reference density, perform normalization processing on the difference data, and uniformly scale it to the standard range to obtain the clustering density normalized difference;

[0034] S513: According to the clustering density normalized difference, arrange the density offset values in the order of the time window center points, merge the time windows within each clustering area and superimpose the normalized results, reconstruct the complete and sequentially consistent sampling sequence structure, and establish the dynamic change dataset of the seedling raising environment.

[0035] The Internet of Things-based forestry seedling raising monitoring system is used to execute the above-mentioned Internet of Things-based forestry seedling raising monitoring method, and the system includes:

[0036] The fluctuation extraction module obtains the soil humidity, temperature, and light intensity data of the monitoring points in the seedling raising environment, detects the adjacent difference reverse points as the fluctuation extreme points, judges the continuous decreasing trend after the fluctuation, and generates the fluctuation boundary discrete distribution map;

[0037] The collaborative calculation module calculates the number of sign changes of the light and humidity difference based on the reverse data sequence in the fluctuation boundary discrete distribution map, aggregates the frequency values, and generates the abnormal fluctuation collaborative frequency data;

[0038] The synchronization evaluation module generates the monitoring point environmental factor synchronization index by identifying the environmental factor coupling strength index according to the abnormal fluctuation collaborative frequency data, aggregating the distribution information by monitoring point, and combining the monitoring point response time distribution;

[0039] The time series analysis module filters out the abnormal monitoring points based on the monitoring point environmental factor synchronization index, extracts the time series and the sampling interval series, calculates the corresponding difference mean value, and generates the environmental factor synchronization deviation distribution map;

[0040] The density normalization module counts the change in the number of sampling points within the time window based on the environmental factor synchronization deviation distribution map, calculates the density gradient difference and performs normalization processing, and generates the dynamic change dataset of the seedling raising environment.

[0041] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0042] In the present invention, by performing difference operations on adjacent time series to locate the extreme points of fluctuations and generating a discrete distribution map, the abnormal fluctuation characteristics of the seedling raising environment parameters are effectively captured, and the real-time performance and accuracy of data anomaly detection are improved. After extracting the abnormal fluctuation sequence based on the discrete distribution map of the fluctuation boundary, the co-variation direction vector of light and humidity is identified and the co-variation frequency data is aggregated and statistically analyzed by segment, so as to quantify the dynamic correlation of environmental factors from multiple dimensions and enhance the granularity of the analysis of the interaction between environmental factors. By extracting the co-variation distribution information of the monitoring points and calculating the coupling intensity index in combination with the response time distribution, an evaluation system for the synchronism of environmental factors is established to accurately identify the coupling relationship between environmental factors at different monitoring points, providing quantitative support for the analysis of the causes of abnormal fluctuations. After screening the monitoring points with synchronism deviation thresholds, a synchronism deviation distribution map is generated by analyzing the corresponding differences between the sampling interval and the response time interval, revealing the spatio-temporal heterogeneity characteristics of the dynamic changes of environmental factors and assisting in optimizing the layout of monitoring points. The difference operation between the sampling density gradient after clustering and recombination and the reference density is combined with normalization processing to construct a dynamic change data set, realizing the continuous mapping of the seedling raising environment parameters in the time and space dimensions, improving the adaptability of the monitoring data to the complex environment dynamics, and breaking through the limitations of traditional single-point static monitoring through multi-dimensional data processing and spatio-temporal correlation modeling. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a schematic diagram of the working process of the present invention;

[0044] Figure 2 It is a flowchart for obtaining the discrete distribution map of the fluctuation boundary in the present invention;

[0045] Figure 3 It is a flowchart for obtaining the co-variation frequency data of abnormal fluctuations in the present invention;

[0046] Figure 4 It is a flowchart for obtaining the synchronism index of environmental factors at the monitoring points in the present invention;

[0047] Figure 5 It is a flowchart for obtaining the synchronism deviation distribution map of environmental factors in the present invention;

[0048] Figure 6 It is a flowchart for obtaining the dynamic change data set of the seedling raising environment in the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0049] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present 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 only used to explain the present invention and are not used to limit the present invention.

[0050] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by terms such as "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present invention. In addition, in the description of the present invention, the meaning of "a plurality of" is two or more, unless otherwise specifically defined.

[0051] Embodiment 1

[0052] Please refer to Figure 1 , the present invention provides a technical solution: a forestry seedling cultivation monitoring method based on the Internet of Things, including the following steps:

[0053] S1: Obtain the soil humidity, temperature and light intensity data of the monitoring points in the seedling cultivation environment, perform adjacent time series difference operations, locate the fluctuation extreme points and determine the starting position of the downward trend, determine the abnormal fluctuation range, and generate a discrete distribution map of the fluctuation boundary;

[0054] S2: Based on the discrete distribution map of the fluctuation boundary, read the abnormal fluctuation sequence, identify the co-variation direction vector of light and humidity, perform co-variation counting on the vector and aggregate it by segment, and count the co-variation frequency of light and humidity to obtain the co-variation frequency data of abnormal fluctuations;

[0055] S3: According to the co-variation frequency data of abnormal fluctuations, extract the co-variation distribution information of the monitoring points, combine the response time distribution of the environmental factors of the monitoring points, identify the coupling intensity index of the environmental factors, and generate the synchronization index of the environmental factors of the monitoring points;

[0056] S4: Based on the synchronization index of the environmental factors of the monitoring points, screen the set of monitoring points whose synchronization deviates from the median threshold, analyze the corresponding differences between the sampling point interval and the response time interval, and generate a synchronization deviation distribution map of the environmental factors;

[0057] S5: According to the synchronization deviation distribution map of the environmental factors, perform clustering analysis on the density distribution of the reorganized sampling points, count the sampling density gradient of the clustering area, perform difference operations with the reference density and perform normalization processing to generate a dynamic change data set of the seedling cultivation environment.

[0058] The discrete distribution map of the fluctuation boundary includes the distribution of the positions of the fluctuation extreme points, the starting points of the downward trend, and the abnormal fluctuation range. The abnormal fluctuation co-frequency data includes the co-variation frequency of the fluctuation segments, the number of co-variation vectors, and the characteristics of the change in the fluctuation direction. The synchronization index of the environmental factors at the monitoring points includes the co-variation distribution information, the response time distribution, and the coupling strength index of the environmental factors. The synchronization deviation map of the environmental factors includes the time distribution of the sampling points, the average value of the sampling point intervals, and the corresponding difference distribution of the monitoring points. The dynamic change data set of the seedling-raising environment includes the time window clustering area, the change rate of the sampling density gradient, and the normalized difference area distribution record.

[0059] Please refer to Figure 2 , and the specific steps for obtaining the discrete distribution map of the fluctuation boundary are as follows:

[0060] S111: Obtain the soil humidity, temperature, and light intensity data of the monitoring points in the seedling-raising environment. Perform adjacent time series difference operations on the continuous data, identify the points where the difference sign changes, and extract the points when the two points before and after the current data point show reverse differences as the fluctuation extreme points to obtain the sequence of the distribution positions of the extreme points;

[0061] Obtain the continuously collected soil humidity data of each monitoring point in the seedling-raising environment over a period of time, for example, record it every 1 hour. The temperature data and light intensity data are also recorded at the same frequency. For the continuous data of each environmental factor, subtract the data of the previous time point from the data of the next time point to obtain a difference sequence. For example, for soil humidity, if the reading at the 2nd hour is 25%RH and the reading at the 1st hour is 23%RH, the difference is +2%RH. Continuously perform this calculation to obtain the humidity difference sequence, temperature difference sequence, and light intensity difference sequence of the entire time series. Check each difference sequence to find the points where the signs of two adjacent differences change. For example, one difference is positive and the following difference is negative, or vice versa. Mark the data points corresponding to the sign change points as the fluctuation extreme points. For soil humidity, if the difference at the 5th hour is +1.5%RH and the difference at the 6th hour is -0.8%RH, then both the data points at the 5th hour and the 6th hour are extracted. Arrange all the extracted fluctuation extreme points according to their positions in the original data sequence. For example, in the 10-hour monitoring of a certain monitoring point, the obtained extreme point positions are the 3rd hour, the 6th hour, and the 8th hour, and the sequence of the distribution positions of the extreme points of this monitoring point is obtained, that is, [3, 6, 8].

[0062] S112: According to the sequence of the distribution positions of the extreme points, for each extreme point and the adjacent data sequence, calculate the continuous difference and judge whether it is less than zero, mark the starting point of the downward trend, and combine the extreme points to form a paired structure to obtain the sequence of the abnormal fluctuation range;

[0063] According to the extreme point distribution position sequence, such as [3, 6, 8], for the first extreme point in the sequence (the 3rd hour) and its adjacent data sequence (for example, the data from the 1st hour to the 5th hour), calculate the difference between each consecutive data point and determine whether the difference is less than zero. If all consecutive differences are less than zero, it indicates that this section of data shows a downward trend. Record the first data point with a difference less than zero as the starting point of the downward trend. For example, if the humidity difference from the 2nd hour to the 3rd hour is -0.5%RH, then the 2nd hour is the starting point of the downward trend. Combine this starting point with the subsequent extreme point (the 3rd hour) to form a paired structure, representing a downward fluctuation range. Repeat this process for each extreme point in the sequence, and also consider the upward trend (the case where consecutive differences are greater than zero, and the starting point is marked as the starting point of the upward trend). Each sequence contains a starting point and an extreme point. For example, [(2, 3), (5, 6), (7, 8)] indicates that there are abnormal fluctuations from the 2nd hour to the 3rd hour, from the 5th hour to the 6th hour, and from the 7th hour to the 8th hour, and obtain the abnormal fluctuation range sequence of this monitoring point.

[0064] S113: Based on the abnormal fluctuation range sequence, group by the starting time of the fluctuation segment, calculate the start - end difference, downward interval, and average downward rate of each group of fluctuation segments, integrate the boundary characteristics of the fluctuation segments, and generate a discrete distribution map of the fluctuation boundary;

[0065] Based on the abnormal fluctuation range sequence, such as [(2, 3), (5, 6), (7, 8)], group according to the starting time of each fluctuation segment. If the starting times of multiple fluctuation segments are the same, they will be grouped together. For each fluctuation segment within each group, calculate the start - end difference of this fluctuation segment (the value of the extreme point minus the value of the starting point). For example, if the humidity at the 2nd hour is 24%RH and the humidity at the 3rd hour is 22%RH, then the start - end difference is 22 - 24 = -2%RH. Calculate the downward interval of this fluctuation segment (the end time minus the starting time). For example, the downward interval from 2 to 3 hours is 3 - 2 = 1 hour. Calculate the average downward rate of this fluctuation segment (the start - end difference divided by the downward interval). For example, -2%RH / 1 hour = -2%RH / hour. Integrate the starting time, end time, start - end difference, downward interval, and average downward rate of each fluctuation segment as the boundary characteristics of this fluctuation segment. Finally, generate a discrete distribution map of the fluctuation boundary. This map can be represented in the form of a table, where each row represents a fluctuation segment and contains information such as its starting time, end time, start - end difference, downward interval, and average downward rate. For example:

[0066] Table 1: Discrete Distribution Table of Fluctuation Boundary

[0067]

[0068] As shown in Table 1, this table records the humidity fluctuation characteristics in different time periods.

[0069] Please refer to Figure 3 , the specific steps for obtaining the co-occurrence frequency data of abnormal fluctuations are as follows:

[0070] S211: Based on the discrete distribution map of the fluctuation boundary, read the abnormal fluctuation subsequences within the corresponding time range, identify the index sequence of each fluctuation in ascending order of time, perform reverse sorting on the light intensity and humidity data sequences, and obtain the light intensity and humidity difference vector sequences;

[0071] Based on the discrete distribution map of the fluctuation boundary, read the abnormal fluctuation subsequences within a specific time range from it. For example, select the fluctuation segment with the starting time between the 2nd hour and the 8th hour, arrange it in ascending order according to its starting time, and identify the index sequence of each fluctuation. This index sequence corresponds to the time points in the original monitoring data. For example, the fluctuation segment (2, 3) corresponds to the index [2, 3], the fluctuation segment (5, 6) corresponds to the index [5, 6], and the fluctuation segment (7, 8) corresponds to the index [7, 8]. For the light intensity data sequence and the humidity data sequence within the same time range, calculate the differences between adjacent data points respectively to obtain the light intensity difference vector sequence and the humidity difference vector sequence, and perform reverse sorting on both the light intensity difference vector sequence and the humidity difference vector sequence to obtain the reverse-sorted light intensity difference vector sequence and the reverse-sorted humidity difference vector sequence.

[0072] S212: According to the light intensity and humidity difference vector sequences, extract the sign changes of adjacent elements in the differences. If it is less than zero, it is regarded as a co-occurrence change event, and count the number of co-occurrence changes in each index segment to obtain the co-occurrence change count sequence;

[0073] According to the reverse-sorted light intensity difference vector sequence and the reverse-sorted humidity difference vector sequence, for the two differences at the same time point, extract the signs. For example, if the light intensity difference at a certain time point is +5 Lux and the humidity difference is -0.3%RH, the signs are positive and negative respectively. Judge whether these two signs are opposite. If one is positive and the other is negative, or one is negative and the other is positive, it is considered that a co-occurrence change event occurs at this time point. Count the number of co-occurrence change events within each fluctuation segment (defined by the starting time and the ending time). For example, for the fluctuation segment (2, 3), if the number of times the sign change of the differences between the corresponding time points 2 to 3 satisfies the co-occurrence change condition is 1 time, the co-occurrence change count of this fluctuation segment is 1. Perform this statistics on each abnormal fluctuation subsequence to obtain the co-occurrence change count sequence. For example, for the fluctuation segments (2, 3), (5, 6), (7, 8), the obtained co-occurrence change count sequence is [1, 0, 2], indicating that the co-occurrence change occurs 1 time in the first fluctuation segment, 0 times in the second fluctuation segment, and 2 times in the third fluctuation segment.

[0074] S213: Calculate the co-variation frequency per unit time within each fluctuation segment according to the co-variation count sequence, combined with the time-span information of each fluctuation segment, and count the corresponding frequencies of the fluctuation segments to establish the co-variation frequency data for abnormal fluctuations.

[0075] According to the co-variation count sequence, such as [1, 0, 2], combined with the time-span information of each fluctuation segment, for example, the time-span of the fluctuation segment (2, 3) is 3 - 2 = 1 hour, calculate the co-variation frequency per unit time within each fluctuation segment. The calculation formula is: co-variation frequency = co-variation count / time-span. For example, the co-variation frequency of the first fluctuation segment is 1 / 1 = 1 time / hour, the second fluctuation segment is 0 / 1 = 0 time / hour, and the third fluctuation segment is 2 / 1 = 1 time / hour. Count the co-variation frequencies corresponding to each fluctuation segment, associate the frequency values with the corresponding fluctuation segment information, and establish the co-variation frequency data for abnormal fluctuations. This data can be presented in tabular form, including information such as the start time, end time, and co-variation frequency of the fluctuation segments. For example:

[0076] Table 2: Data Table of Co-variation Frequency for Abnormal Fluctuations

[0077]

[0078] As shown in Table 2, this table shows the co-variation frequencies within different fluctuation segments.

[0079] Please refer to Figure 4 , and the specific steps for obtaining the synchrony index of the monitoring point environmental factors are as follows:

[0080] S311: Aggregate according to the monitoring point number mapping based on the co-variation frequency data for abnormal fluctuations, extract the associated abnormal fluctuation numbers and co-variation frequencies, and calculate the co-variation distribution of each monitoring point in abnormal fluctuations by grouping to generate the co-variation distribution sequence of the monitoring points.

[0081] Based on the co-variation frequency data for abnormal fluctuations, aggregate according to the monitoring point numbers. If multiple monitoring points have abnormal fluctuations within the same time period, then associate the abnormal fluctuation numbers and the corresponding co-variation frequencies, extract the abnormal fluctuation numbers associated with each monitoring point and their co-variation frequencies. For example, the abnormal fluctuation numbers of monitoring point A are [W1, W3], and the corresponding co-variation frequencies are [1.5, 0.8] times / hour. Group according to the monitoring point numbers and calculate the co-variation distribution of each monitoring point in all abnormal fluctuations. For example, statistical quantities such as the average co-variation frequency and the variance of the co-variation frequency of each monitoring point can be calculated to generate the co-variation distribution sequence of the monitoring points. This sequence contains the numbers of each monitoring point and their corresponding co-variation distribution characteristics. For example, the co-variation distribution sequence of monitoring point A is expressed as its average co-variation frequency of 1.15 times / hour.

[0082] S312: Based on the co-varying distribution sequence of monitoring points, extract the response time series of monitoring points, calculate the fluctuations corresponding to the distribution information at each time point, evaluate the degree of time distribution fluctuations of each monitoring point's response sequence, and generate the response time distribution sequence of monitoring points;

[0083] Based on the co-varying distribution sequence of monitoring points, extract the response time series when abnormal fluctuations occur at each monitoring point. The response time can be defined as the time interval from the start of obvious fluctuations in environmental factors to the start of co-varying events. For example, if the humidity at monitoring point A drops at the 2nd hour and the co-variation of light and humidity starts at the 2.5th hour, then its response time is 0.5 hour. Calculate the response time of each monitoring point in different abnormal fluctuation events to obtain the response time series of monitoring points. For example, the response time series of monitoring point A is [0.5, 0.3, 0.7] hours. Calculate the degree of fluctuations of each monitoring point's response sequence corresponding to its distribution information, which can be measured using the standard deviation. The larger the standard deviation, the more dispersed the response time distribution and the greater the degree of fluctuations. Generate the response time distribution sequence of monitoring points, which includes the number of each monitoring point and the statistical characteristics (such as the standard deviation) of its response time distribution.

[0084] S313: According to the response time distribution sequence of monitoring points, combined with the short-term environmental factor distribution of monitoring points, statistically calculate the mean value, energy concentration degree, and standard deviation, evaluate the coupling structure of environmental factors, and calculate the coupling strength index using the formula:

[0085] ;

[0086] Obtain the environmental factor synchronization index of monitoring points;

[0087] where, represents the environmental factor synchronization index of monitoring points, represents the humidity data of the th monitoring point, represents the mean value of the humidity data of the th monitoring point, represents the light intensity data of the th monitoring point, represents the response time of the th monitoring point, represents the total number of monitoring points;

[0088] Based on the response time distribution sequence of the monitoring points, combined with the distribution of environmental factors at each monitoring point within a short period of time, for example, the mean, energy concentration, and standard deviation of soil moisture, temperature, and light intensity within the past 1 hour, where the energy concentration can be expressed as the average of the sum of the squares of the values of each environmental factor, statistical statistics are calculated to evaluate the coupling structure between different environmental factors. For example, if the change trends of humidity and temperature are consistent and the response times are similar, it is considered that the coupling is strong. The coupling strength index is calculated. For example, the Pearson correlation coefficient can be used to measure the correlation between pairwise environmental factors, and the correlation coefficients are synthesized to obtain an overall coupling strength index. Finally, the environmental factor synchronization index of each monitoring point is obtained, which reflects the degree of co-variation and response synchronization between different environmental factors at this monitoring point;

[0089] The environmental factor synchronization index of the monitoring point reflects the degree of co-variation and response synchronization between different environmental factors at the same monitoring point. This index evaluates the coupling degree by measuring the fluctuations of environmental factors such as humidity and light intensity within a specific time and their relationships. The lower the value of the synchronization index, the worse the synchronization between different environmental factors, indicating a large degree of inconsistent changes or asynchronous responses. While the higher the value of the synchronization index, it shows that the environmental factors have a strong correlation or coordination in response, demonstrating good synchronization. This index can be used to monitor and optimize the interaction of environmental factors, and thus help to more accurately predict environmental changes and their impacts in fields such as agriculture and meteorology;

[0090] : The environmental factor synchronization index of the monitoring point, which measures the degree of co-variation and response synchronization between the environmental factors of the monitoring point;

[0091] : The humidity data of the

[0092] th monitoring point, with the unit of volumetric water content (m³ / m³), measured using an SM150T soil moisture sensor. After specific calibration for the soil, the accuracy of humidity measurement is ±3% (m³ / m³);

[0093] : The mean value of the humidity data of the

[0094] [[ID=۲۸]] : The response time of the th monitoring point, in seconds (s). According to the technical specifications of the sensor, the response time of the SM150T soil moisture sensor is 1 second;

[0095] : The total number of all monitoring points;

[0096] To ensure the dimensional consistency of each parameter in the formula, all data have been converted to the same unit:

[0097] The humidity data ( and ) are expressed in volumetric water content (m³ / m³), the light intensity data ( ) are expressed in lux, and the response time ( ) is expressed in seconds (s);

[0098] Suppose there are 3 monitoring points ( ), and the corresponding humidity data, light intensity data, and response time are as follows:

[0099] ;

[0100] ;

[0101] ;

[0102] ;

[0103] The calculation steps are as follows:

[0104] Calculate :

[0105] ;

[0106] ;

[0107] ;

[0108] Calculate : ;

[0109] Calculate :

[0110] Suppose is the mean value of the light intensity data:

[0111] ;

[0112] ;

[0113] ;

[0114] ;

[0115] Calculate : ;

[0116] Calculate : ;

[0117] Calculate : ;

[0118] The calculated synchronization index represents the degree of co-variation and response synchronization among environmental factors at the monitoring points. A lower value means poorer synchronization among environmental factors, and further analysis and adjustment are required.

[0119] Please refer to Figure 5 , and the specific steps for obtaining the distribution map of environmental factor synchronization deviation are as follows:

[0120] S411: Based on the environmental factor synchronization index at the monitoring points, identify the set of synchronization values according to the monitoring point numbers, extract the quartiles and set the abnormal deviation threshold, screen the monitoring point index values, and output the set of abnormal monitoring point numbers;

[0121] Based on the environmental factor synchronization index at the monitoring points, for example, the synchronization index values of a group of monitoring points are [0.8, 0.5, 0.9, 0.3, 0.7]. Identify and aggregate them according to the monitoring point numbers, extract the set of synchronization values corresponding to each monitoring point, and calculate the quartiles of this set, including the lower quartile (Q1), the median (Q2), and the upper quartile (Q3). For example, for the above data, after sorting, it is [0.3, 0.5, 0.7, 0.8, 0.9], then Q1 = 0.4, Q2 = 0.7, Q3 = 0.85. Set the abnormal deviation threshold, which can be determined based on the interquartile range (IQR = Q3 - Q1). For example, values lower than Q1 - 1.5 * IQR or higher than Q3 + 1.5 * IQR can be regarded as abnormal deviations. Calculate IQR = 0.85 - 0.4 = 0.45, then the lower limit is 0.4 - 1.5 * 0.45 = -0.275, and the upper limit is 0.85 + 1.5 * 0.45 = 1.525. Screen out the monitoring points whose synchronization index values exceed this threshold. For example, if the synchronization index of a certain monitoring point is 0.3, it is considered an abnormal deviation, and output the set of abnormal monitoring point numbers, for example, {M4}.

[0122] S412: According to the set of abnormal monitoring point numbers, extract the interval differences between consecutive sampling points of the monitoring points and standardize them, calculate the ratio of the corresponding time of the fluctuation segment to the sampling misalignment, aggregate and find the average value of the misalignment ratio sequences of each monitoring point to generate the average sequence of alignment deviation of the monitoring points;

[0123] According to the set of abnormal monitoring point numbers, such as {M4}, extract the time interval differences between consecutive sampling points of the abnormal monitoring points. For example, if the sampling timestamps of monitoring point M4 are [1:00, 1:30, 2:00, 2:45], the interval differences between consecutive sampling points are [30 minutes, 30 minutes, 45 minutes]. Standardize the interval differences. For example, each interval difference can be divided by the average value of all sampling interval differences to obtain the standardized interval differences. Calculate the ratio of the corresponding time (the duration of the fluctuation segment) of each fluctuation segment to the sampling misalignment (the standard deviation of the sampling interval) of the monitoring point within this fluctuation segment. For example, if a certain fluctuation segment lasts for 1 hour, the sampling intervals of monitoring point M4 within this time period are [30, 45] minutes, the average value is 37.5 minutes, and the standard deviation is approximately 7.5 minutes, then the misalignment ratio is 7.5 / 60≈0.125. Aggregate and find the average value of the misalignment ratio sequences of all fluctuation segments of each monitoring point to obtain the average sequence of alignment deviation of each monitoring point. For example, the average alignment deviation of monitoring point M4 is 0.15.

[0124] S413: Based on the average sequence of alignment deviation of the monitoring points, sort the average alignment deviations according to the monitoring point numbers, plot the synchronization deviation trajectory diagram with the time difference as the ordinate and the monitoring point numbers as the abscissa, and use the formula:

[0125] ;

[0126] Calculate the corresponding deviation between the sampling time and interval of the monitoring points, use the deviation as the discrete trajectory points of the monitoring points, and generate the synchronization deviation distribution diagram of environmental factors;

[0127] where, represents the corresponding deviation between the sampling time and interval of the monitoring point, represents the actual sampling time of the i-th monitoring point, represents the reference time, represents the sampling interval time of monitoring point i, represents the total number of monitoring points;

[0128] Based on the mean sequence of alignment deviations at monitoring points, for example, the monitoring point numbers are [M1, M2, M3, M4], and the corresponding average alignment deviations are [0.05, 0.1, 0.08, 0.15]. Taking the average alignment deviation as the ordinate and the monitoring point number as the abscissa in the order of the monitoring point numbers, a synchronous deviation trajectory graph is plotted. This graph shows the magnitude of the average alignment deviation of each monitoring point in the form of discrete points. Calculate the deviation of the sampling time of each monitoring point corresponding to the ideal equally spaced sampling. For example, if the ideal sampling interval is 30 minutes, and the actual sampling interval of monitoring point M4 is [30, 45] minutes, then its deviation is [0, 15] minutes. Calculate the ratio of the time deviation to the ideal sampling interval, for example, [0 / 30, 15 / 30] = [0, 0.5]. Plot the corresponding deviation as the discrete trajectory point of this monitoring point in the coordinate system with the time difference as the ordinate and the monitoring point number as the abscissa. Finally, an environmental factor synchronous deviation distribution graph is generated. This graph intuitively shows the synchronous deviation and its distribution of each monitoring point in the sampling time;

[0129] The corresponding deviation between the sampling time and interval of a monitoring point refers to the degree of difference between the actual sampling time and the predetermined sampling time of the monitoring point within a certain time range, taking into account the influence of the sampling interval. Specifically, it measures the magnitude of the deviation between the sampling time and the reference time corresponding to the sampling interval. By calculating the corresponding deviation, the time synchronization of monitoring at different sampling points and the stability of the sampling intervals at each sampling point can be evaluated. A larger corresponding deviation indicates a larger deviation in the sampling time, which affects the accuracy of the monitoring data, while a smaller corresponding deviation indicates better system synchronization and more consistent sampling times;

[0130] : The actual sampling time of the th monitoring point, obtained from the timestamp recorded by the monitoring device;

[0131] : The reference time, which is the predetermined standard sampling time, determined by the configuration file or experimental design;

[0132] : The sampling interval time of the th monitoring point, obtained from the sampling frequency of the device or the configuration file;

[0133] : The total number of monitoring points, determined according to the experimental design or device configuration;

[0134] Dimensionality unification process (normalization): All time units are unified to seconds (s). If the unit of the sampling interval time is milliseconds (ms), it needs to be converted to seconds: ;

[0135] Assume there are 3 monitoring points and the reference time seconds, and the sampling interval times are 10, 20, and 30 seconds respectively. The corresponding actual sampling times are 105, 130, and 160 seconds. Then:

[0136] seconds, seconds, seconds;

[0137] seconds, seconds, seconds;

[0138] Calculate the deviation: ;

[0139] The calculated deviation seconds, which represents the mean absolute deviation between the actual sampling time of the monitoring point and the reference time, taking into account the weights of the sampling interval times of each sampling point. This result is used to evaluate the synchronization performance of the monitoring system. The smaller the value, the better the synchronization.

[0140] Please refer to Figure 6 , and the specific steps for obtaining the dynamic change dataset of the seedling-raising environment are as follows:

[0141] S511: According to the environmental factor synchronization deviation distribution map, reorder and sort all sampling points on the time axis, calculate the change trend of the sampling point density, and statistically calculate the change rate of the sampling points per unit time to obtain the sliding window sampling density gradient;

[0142] According to the environmental factor synchronization deviation distribution map, extract the timestamps of all sampling points of all monitoring points, reorder and sort the sampling points on the time axis to obtain a global time series, calculate the time intervals between adjacent sampling points, analyze the change trend of the sampling point density. For example, the density can be reflected by calculating the number of sampling points per unit time (such as 1 hour), and statistically calculate the change rate of the number of sampling points per unit time within consecutive time windows (for example, a sliding window size of 3 hours). The calculation formula is: sampling point change rate = (the number of sampling points in the current time window - the number of sampling points in the previous time window) / the time span of the time window, to obtain the sliding window sampling density gradient sequence. This sequence reflects the change trend and rate of the sampling density over time. For example, if there are 60 sampling points in the previous 3-hour window and 75 sampling points in the current 3-hour window, then the sampling point change rate is (75 - 60) / 3 = 5 points / hour.

[0143] S512: Based on the sliding window sampling density gradient, extract the sampling density within the time window and perform a difference operation with the reference density, and perform a normalization process on the difference data to uniformly scale it to the standard range to obtain the clustering density normalization difference;

[0144] Based on the sliding window sampling density gradient sequence, for each time window, extract the sampling density within that time window (e.g., the number of sampling points per unit time), and perform a difference operation with a reference density, which can be the average sampling density of the entire time series. For example, if the sampling density within a certain time window is 25 points per hour and the reference density is 20 points per hour, then the difference is 25 - 20 = 5 points per hour. Perform normalization processing on the obtained difference data to uniformly scale all differences to a standard range, such as [-1, 1]. Common normalization methods include min-max normalization or Z-score standardization. Assuming min-max normalization is used, if the minimum value of the difference is -5 and the maximum value is 10, then the normalized value is (5 - (-5)) / (10 - (-5))*(1 - (-1)) + (-1) = 10 / 15*2 - 1 ≈ 0.33, obtaining a normalized difference sequence of clustering density.

[0145] S513: Arrange the density offset values according to the center points of the time windows based on the normalized difference of clustering density, aggregate the time windows within each clustering region, and superimpose the normalized results to reconstruct a complete and sequentially consistent sampling sequence structure, establishing a dynamic change dataset of the seedling raising environment;

[0146] According to the normalized difference sequence of clustering density, arrange the density offset values in the time order of the center points of each time window. For example, if the center points of the time windows are the 1st hour, the 4th hour, and the 7th hour, and the corresponding normalized density offset values are [0.2, -0.5, 0.8], then the arranged sequence is [(1, 0.2), (4, -0.5), (7, 0.8)]. Aggregate the time windows within each predefined clustering region. For example, time windows that are adjacent in time and have similar density offset values can be divided into a clustering region, and then the normalized results within each clustering region are superimposed. For example, if the normalized offset values of two adjacent time windows are 0.3 and 0.4 respectively, then the superimposed result is 0.7. In this way, reconstruct a complete and sequentially consistent sampling sequence structure, which reflects the overall offset of the sampling density in different time regions corresponding to the reference density. Finally, establish a dynamic change dataset of the seedling raising environment, which contains time information and the normalized density offset values within the corresponding time periods, and can be used to analyze the dynamic change patterns of data sampling in the seedling raising environment.

[0147] The forestry seedling raising monitoring system based on the Internet of Things is used to execute the above-mentioned forestry seedling raising monitoring method based on the Internet of Things. The system includes:

[0148] The fluctuation extraction module obtains the soil humidity, temperature, and light intensity data of the monitoring points in the seedling raising environment, detects the adjacent difference reversal points as the fluctuation extreme points, judges the continuous decreasing trend after the fluctuation, and generates a discrete distribution map of the fluctuation boundary;

[0149] The collaborative calculation module calculates the number of sign changes in the difference between light and humidity based on the reverse data sequence in the discrete distribution map of the fluctuation boundary, aggregates the frequency values, and generates the collaborative frequency data of abnormal fluctuations;

[0150] The synchronous evaluation module, according to the collaborative frequency data of abnormal fluctuations, aggregates the distribution information by monitoring points, combines the response time distribution of the monitoring points to identify the coupling intensity index of environmental factors, and generates the synchronous index of environmental factors for the monitoring points;

[0151] The time series analysis module, based on the synchronous index of environmental factors for the monitoring points, screens out abnormal monitoring points, extracts the time series and the sampling interval sequence, calculates the corresponding difference mean value, and generates a distribution map of the synchronous deviation of environmental factors;

[0152] The density normalization module, based on the distribution map of the synchronous deviation of environmental factors, counts the change in the number of sampling points within the time window, calculates the density gradient difference and performs normalization processing, and generates a dynamic change data set for the seedling raising environment.

[0153] The above are only the preferred embodiments of the present invention, and do not limit the present invention in other forms. Any person skilled in the art may use the disclosed technical content to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical content of the technical solution of the present invention, any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A forestry seedling cultivation monitoring method based on the Internet of Things, characterized in that, It includes the following steps: S1: Obtain the soil humidity, temperature, and light intensity data of the monitoring points in the seedling raising environment, perform adjacent time series difference operations, locate the fluctuation extreme points, determine the starting position of the downward trend, determine the abnormal fluctuation range, and generate a discrete distribution map of the fluctuation boundary; S2: Based on the discrete distribution map of the fluctuation boundary, read the abnormal fluctuation sequence, identify the co-variation direction vector of light and humidity, perform co-variation counting on the vector and aggregate it by segment, and count the co-variation frequency of light and humidity to obtain the co-variation frequency data of abnormal fluctuations; S3: According to the co-variation frequency data of abnormal fluctuations, extract the co-variation distribution information of the monitoring points, combine the response time distribution of the environmental factors of the monitoring points, identify the coupling intensity index of the environmental factors, and generate the synchronization index of the environmental factors of the monitoring points; S4: Based on the synchronization index of the environmental factors of the monitoring points, screen the set of monitoring points whose synchronization deviates from the median threshold, analyze the corresponding differences between the sampling point intervals and the response time intervals, and generate a distribution map of the synchronization deviation of the environmental factors.

2. The forestry seedling cultivation monitoring method based on the Internet of Things according to claim 1, wherein, The discrete distribution map of the fluctuation boundary includes the distribution of the positions of the fluctuation extreme points, the distribution of the starting points of the downward trend, and the abnormal fluctuation range. The co-variation frequency data of abnormal fluctuations includes the co-variation frequency of the fluctuation segments, the number of co-variations of the vectors, and the change characteristics of the fluctuation direction. The synchronization index of the environmental factors of the monitoring points includes the co-variation distribution information, the response time distribution, and the coupling intensity index of the environmental factors. The distribution map of the synchronization deviation of the environmental factors includes the time distribution of the sampling points, the average value of the sampling point intervals, and the corresponding difference distribution of the monitoring points.

3. The forestry seedling cultivation monitoring method based on the Internet of Things according to claim 1, wherein, The specific steps for obtaining the discrete distribution map of the fluctuation boundary are as follows: S111: Obtain the soil humidity, temperature, and light intensity data of the monitoring points in the seedling raising environment, perform adjacent time series difference operations on the continuous data, identify the points where the difference changes sign, and extract the points when the two points before and after the current data point have reverse differences as the fluctuation extreme points to obtain the sequence of the distribution positions of the extreme points; S112: According to the sequence of the distribution positions of the extreme points, for each extreme point and the adjacent data sequence, calculate the continuous differences and determine whether they are less than zero, mark the starting point of the downward trend, and form a paired structure with the extreme points to obtain the sequence of the abnormal fluctuation ranges; S113: Based on the sequence of the abnormal fluctuation ranges, group them by the starting time of the fluctuation segments, calculate the start and end differences, the downward interval, and the average downward rate of each group of fluctuation segments, integrate the boundary characteristics of the fluctuation segments, and generate a discrete distribution map of the fluctuation boundary.

4. The method for monitoring forestry seedling cultivation based on the Internet of Things according to claim 3, wherein The specific steps for obtaining the co-variation frequency data of abnormal fluctuations are as follows: S211: Based on the discrete distribution map of the fluctuation boundary, read the abnormal fluctuation subsequence within the corresponding time range, identify the index sequence of each segment of the fluctuation in ascending order of time, and perform reverse sorting on the light and humidity data sequences to obtain the light and humidity difference vector sequence; S212: According to the light and humidity difference vector sequence, extract the sign changes of the adjacent elements in the differences. If it is less than zero, it is regarded as a co-variation event, and count the number of co-variations in each segment according to the index segments to obtain the co-variation counting sequence; S213: According to the co-variation count sequence, combined with the time-span information of each fluctuation segment, calculate the co-variation frequency per unit time within each fluctuation segment, count the corresponding frequencies of the fluctuation segments, and establish the co-variation frequency data for abnormal fluctuations.

5. The method for monitoring forestry seedling cultivation based on the Internet of Things according to claim 4, wherein The specific steps for obtaining the synchrony index of the environmental factors at the monitoring points are as follows: S311: According to the co-variation frequency data for abnormal fluctuations, map and aggregate by monitoring point number, extract the associated abnormal fluctuation numbers and co-variation frequencies, calculate the co-variation distribution of each monitoring point in the abnormal fluctuations by grouping, and generate the co-variation distribution sequence of the monitoring points. S312: Based on the co-variation distribution sequence of the monitoring points, extract the response time sequence of the monitoring points, calculate the fluctuations corresponding to the time points in the distribution information, evaluate the degree of time distribution fluctuations of each monitoring point's response sequence, and generate the response time distribution sequence of the monitoring points. S313: According to the response time distribution sequence of the monitoring points, combined with the short-term environmental factor distribution of the monitoring points, statistically calculate the mean value, energy concentration degree, and standard deviation, evaluate the coupling structure of the environmental factors, and calculate the coupling intensity index to obtain the synchrony index of the environmental factors at the monitoring points.

6. The forestry seedling cultivation monitoring method based on the Internet of Things according to claim 5, characterized in that, The specific steps for obtaining the synchrony deviation distribution map of the environmental factors are as follows: S411: Based on the synchrony index of the environmental factors at the monitoring points, identify the set of synchrony values according to the monitoring point numbers, extract the quartiles and set the abnormal deviation threshold, screen the monitoring point index values, and output the set of abnormal monitoring point numbers. S412: According to the set of abnormal monitoring point numbers, extract the difference between consecutive sampling point intervals of the monitoring points and standardize it, calculate the ratio of the corresponding time of the fluctuation segment to the sampling misalignment, aggregate and calculate the average value of the misalignment ratio sequence of each monitoring point, and generate the average misalignment deviation sequence of the monitoring points. S413: Based on the average misalignment deviation sequence of the monitoring points, sort the average misalignment deviations according to the monitoring point numbers, plot the synchrony deviation trajectory map with the time difference as the vertical coordinate and the monitoring point numbers as the horizontal coordinate, calculate the corresponding deviation between the sampling time and interval of the monitoring points, and use the deviation as the discrete trajectory points of the monitoring points to generate the synchrony deviation distribution map of the environmental factors.

7. The method for monitoring forestry seedling cultivation based on the Internet of Things according to claim 1, wherein The method further includes step S5: S5: According to the synchrony deviation distribution map of the environmental factors, perform clustering analysis on the density distribution of the reorganized sampling points, statistically calculate the sampling density gradient in the clustering region, perform difference operation with the reference density and perform normalization processing to generate the dynamic change dataset of the seedling-raising environment. The dynamic change dataset of the seedling-raising environment includes the time-window clustering region, the change rate of the sampling density gradient, and the normalized difference region distribution record.

8. The method for monitoring forestry seedling cultivation based on the Internet of Things according to claim 7, characterized in that, The specific steps for obtaining the dynamic change dataset of the seedling-raising environment are as follows: S511: According to the synchrony deviation distribution map of the environmental factors, reorganize and sort all the sampling points on the time axis, calculate the change trend of the sampling point density, and statistically calculate the change rate of the sampling points per unit time to obtain the sliding-window sampling density gradient. S512: Based on the sliding-window sampling density gradient, extract the sampling density within the time window and perform difference operation with the reference density, perform normalization processing on the difference data, and uniformly scale it to the standard range to obtain the normalized difference of the clustering density. S513: Arrange the density offset values in the order of the center points of the time windows according to the normalized difference of the clustering density, merge the time windows within each clustering region and superimpose the normalized results to reconstruct a complete and sequentially consistent sampling sequence structure, and establish a dynamic change dataset of the seedling raising environment.

9. A forestry seedling cultivation monitoring system based on the Internet of Things, characterized in that, The system is used to implement the Internet of Things-based forestry seedling raising monitoring method according to any one of claims 1-8. The system includes: The fluctuation extraction module obtains the soil humidity, temperature and light intensity data of the monitoring points in the seedling raising environment, detects the adjacent difference reverse points as the fluctuation extreme points, judges the continuous decreasing trend after the fluctuation, and generates a discrete distribution map of the fluctuation boundary. The collaborative calculation module calculates the number of sign changes of the difference between light and humidity based on the reverse data sequence in the discrete distribution map of the fluctuation boundary, aggregates the frequency values, and generates abnormal fluctuation collaborative frequency data. The synchronous evaluation module generates a monitoring point environmental factor synchronization index by identifying the coupling strength index of environmental factors according to the abnormal fluctuation collaborative frequency data, aggregating the distribution information by monitoring points, and combining the response time distribution of the monitoring points. The time series analysis module filters out abnormal monitoring points based on the monitoring point environmental factor synchronization index, extracts the time series and the sampling interval sequence, calculates the corresponding difference mean value, and generates a synchronous deviation distribution map of environmental factors. The density normalization module counts the change in the number of sampling points within the time window based on the synchronous deviation distribution map of environmental factors, calculates the density gradient difference and performs normalization processing to generate a dynamic change dataset of the seedling raising environment.

Images