CN-121998436-A - Full-flow digital early warning and blockchain responsibility tracing method for tea

CN121998436ACN 121998436 ACN121998436 ACN 121998436ACN-121998436-A

Abstract

The invention relates to the technical field of multi-source remote sensing data processing, and discloses a full-flow digital early warning and blockchain responsibility tracing method for tea, which comprises the steps of firstly constructing a terraced field planting micro-unit distribution map and analyzing repeated distances of terraced field landforms; and then acquiring multi-source heterogeneous sensing data, determining comprehensive spatial resolution and comprehensive time acquisition interval, and calculating a space-time sampling aliasing index according to the comprehensive spatial resolution and the comprehensive time acquisition interval. And generating a monitoring credibility attenuation factor by using the index, and recovering an anti-aliasing signal as a regular constraint to obtain a credibility correction risk value. The method generates an early warning event according to the corrected risk value, calculates responsibility tracing and apportioning weights by using the monitoring credibility attenuation factors, and finally stores the early warning event, the responsibility weights and the recovery parameters into a blockchain system. The invention effectively solves the problem of observation distortion caused by cloud and fog shielding and terrain crushing in mountain tea gardens, and ensures the accuracy and reliability of risk early warning and responsibility identification.

Inventors

LI GUOSHI
ZHANG HAIJIN
Zhou Huixu
SHAO ZHENHUA

Assignees

南平市计量所
闽江大学

Dates

Publication Date: 20260508
Application Date: 20260409

Claims (9)

1. The whole-flow digital early warning and blockchain responsibility tracing method for the tea is characterized by comprising the following steps of: Constructing a terrace planting micro-unit distribution map, and analyzing terrace landform repetition intervals of each terrace planting micro-unit along the slope direction; acquiring multi-source heterogeneous sensing data aiming at the terraced field planting micro unit, setting a preset time window, calculating and determining comprehensive spatial resolution and comprehensive time acquisition interval in the preset time window, and acquiring continuous aging of a risk process of the terraced field planting micro unit; calculating to obtain a space-time sampling aliasing index based on the ratio relation between the comprehensive spatial resolution and the terrace landform repetition interval and the ratio relation between the comprehensive time acquisition interval and the continuous aging of the risk process; Generating a monitoring credibility attenuation factor by using the space-time sampling aliasing index, and executing anti-aliasing signal restoration processing on the multi-source heterogeneous sensing data by taking the monitoring credibility attenuation factor as a constraint condition to obtain restoration model parameters and credibility correction risk values; and generating an early warning event according to the credibility correction risk value, calculating responsibility tracing and apportionment weights of the corresponding tea batches according to the monitoring credibility attenuation factors, and storing the early warning event, the responsibility tracing and apportionment weights and the restoration model parameters into a blockchain system for tracing.
2. The full-flow digital early warning and blockchain responsibility tracing method for tea leaves of claim 1, wherein constructing a terrace planting microcell distribution map comprises: collecting a digital elevation model and a high-resolution orthophoto of a tea garden area, and identifying a slope dominant direction vector based on the digital elevation model; for any spatial point in the tea garden, projecting coordinates of the spatial point onto the slope leading direction vector, and calculating to obtain slope projection coordinates; setting a fixed micro-unit segmentation length, carrying out equidistant segmentation on a tea garden area along the axial direction of the slope leading direction vector by utilizing the slope projection coordinates to generate a plurality of discrete terrace planting micro-unit sets, and forming the terrace planting micro-unit distribution map based on the plurality of discrete terrace planting micro-unit sets.
3. The full-flow digital early warning and blockchain responsibility tracing method for tea leaves according to claim 1, wherein analyzing terrace land feature repetition intervals of each terrace planting micro unit along a slope direction comprises: sampling along the slope profile of each terrace planting micro unit to obtain an elevation sequence; Calculating a second-order difference based on the high program sequence, and constructing a slope curvature sequence; performing autocorrelation analysis on the slope curvature sequence, and calculating a curvature autocorrelation function; And identifying a first obvious non-zero peak value of the curvature autocorrelation function in a preset searching range, and taking a hysteresis distance corresponding to the first obvious non-zero peak value as the terrace landform repetition interval.
4. The method for full-process digital early warning and blockchain responsibility tracing of tea leaves according to claim 1, wherein the calculating and determining the comprehensive spatial resolution and the comprehensive time acquisition interval in the predetermined time window comprises the following steps: Acquiring multiple types of data sources aiming at the terraced fields planting micro unit in the preset time window, acquiring single-source spatial resolution of each type of data source and setting a corresponding source weight coefficient; Calculating the square reciprocal of each single-source spatial resolution, carrying out weighted summation on the square reciprocal by utilizing the source weight coefficient, and squaring the reciprocal of the weighted summation to obtain the comprehensive spatial resolution; acquiring effective observation time stamp sequences arranged in time sequence in the preset time window, and calculating a time difference value set between two adjacent effective observation time stamp sequences; and calculating the median of the time difference value set, and taking the median as the comprehensive time acquisition interval.
5. The method for full-process digital early warning and blockchain responsibility tracing of tea leaves according to claim 1, wherein the risk process of obtaining the terrace planting micro-unit is continuous in time and the time-space sampling aliasing index is calculated, comprising: Acquiring a microclimate time sequence of the terrace planting microcell, performing mean value removal treatment, and calculating a normalized autocorrelation function; Searching the minimum hysteresis step number when the normalized autocorrelation function is attenuated to be lower than a preset attenuation threshold value, and taking the product of the minimum hysteresis step number and the comprehensive time acquisition interval as the continuous aging of the risk process; Calculating a first ratio of twice the comprehensive spatial resolution to the terrace topography repetition interval, and calculating a second ratio of twice the comprehensive time acquisition interval to the continuous aging of the risk process; and selecting the maximum value of the first ratio and the second ratio as the space-time sampling aliasing index.
6. The method for full-process digital early warning and blockchain responsibility tracing of tea leaves according to claim 1, wherein generating a monitoring confidence attenuation factor by using the space-time sampling aliasing index comprises: Setting an attenuation adjustment constant, and calculating a difference value between the space-time sampling aliasing index and a numerical value I; and calculating the product of the attenuation adjustment constant and the difference value, taking the product as an index to calculate a natural index function value, and taking the natural index function value as the monitoring credibility attenuation factor.
7. The method for full-process digital early warning and blockchain responsibility tracing of tea leaves according to claim 1, wherein the steps of performing anti-aliasing signal restoration processing on the multi-source heterogeneous sensing data by using the monitoring confidence attenuation factor as a constraint condition to obtain restoration model parameters and confidence correction risk values include: Constructing a signal recovery basis function model containing a slowly-varying background term and a periodic harmonic term based on the terrace landform repetition interval; Constructing an optimization objective function, wherein the optimization objective function comprises a data fitting error term and a smooth regular term taking the monitoring credibility attenuation factor as a weight; Solving restoration model parameters which minimize the optimization objective function, and reconstructing by using the restoration model parameters to obtain restoration monitoring signals of the terrace planting micro units; and calculating the fluctuation energy of the restoration monitoring signal in the preset time window, calculating the ratio of the fluctuation energy to the monitoring reliability attenuation factor, and taking the ratio as the reliability correction risk value.
8. The full-process digital early warning and blockchain responsibility tracing method of the tea according to claim 1, wherein the steps of generating early warning events according to the credibility correction risk values and calculating responsibility tracing and sharing weights of corresponding tea batches according to the monitoring credibility attenuation factors include: Constructing an early warning probability logic function comprising a linear term of the reliability correction risk value and a logarithmic term of the space-time sampling aliasing index, calculating early warning probability, and generating the early warning event based on the early warning probability; acquiring the contribution of the terrace planting micro unit to the specific tea batch, and calculating the ratio of the contribution to the monitoring credibility attenuation factor to obtain intermediate weight; And carrying out normalization processing on the intermediate weights of all terrace planting micro units belonging to the same tea batch to obtain the responsibility tracing and sharing weight of each terrace planting micro unit aiming at the tea batch.
9. The method of claim 8, wherein storing the early warning event, the responsibility tracing apportionment weights, and the recovery model parameters into a blockchain system for tracing comprises: Constructing a double account book storage structure comprising a fast account book and an evidence account book; extracting an event identifier, a time stamp and a corresponding event load of the early warning event, and extracting a batch identifier of the specific tea batch; writing the event identification, the batch identification, the time stamp, the space-time sampling aliasing index, the early warning probability and the hash fingerprint of the event load into the quick ledger; And writing the multi-source heterogeneous perception data serving as original data into a distributed file system, acquiring a content addressing identifier, and writing the content addressing identifier, the hash fingerprint of the restoration model parameter and the monitoring credibility attenuation factor into the evidence ledger.

Description

Full-flow digital early warning and blockchain responsibility tracing method for tea Technical Field The invention relates to the technical field of multi-source remote sensing data processing, in particular to a full-flow digital early warning and blockchain responsibility tracing method for tea. Background In the current digitization process of tea industry, satellite remote sensing, unmanned aerial vehicle images and ground internet of things sensors are utilized to cooperatively monitor a tea garden, and product tracing is realized by combining a blockchain technology, so that the technology becomes an important means for guaranteeing tea quality and safety. In the prior art, a monitoring area is generally regarded as a continuous smooth surface or a regular grid, environmental information such as spectrum data, temperature and humidity parameters and the like is directly collected, the data are used as a basis for judging production risks (such as plant diseases and insect pests and frost), and then related data are recorded in a non-tamperable way through a blockchain to clear production responsibility. However, in actual mountain and hilly tea areas, the physical form and production environment of tea gardens have extremely strong specificity, so that obvious systematic deviation exists in the existing monitoring and tracing system: First, mountain tea garden is not a continuous natural slope, but a terraced field composite structure constructed by engineering means. The structure is formed by a series of nearly horizontal terraces alternating with the steeply pitched (or walls) supporting the terraces, forming micro-topographical elements that repeat periodically along the slope. Different micro-topography units (such as terrains, terraces, interlines) have significant meter-level differences in soil physicochemical properties, microbial community distribution and disease occurrence probability. Second, existing monitoring means often have difficulty matching the inherent structure of the mountain environment in both the spatial and temporal dimensions. In the space dimension, the space resolution (typically 10 meters) of the mainstream monitoring means such as satellite remote sensing is far lower than the repeated scale (typically 1-3 meters) of the terraced field micro-topography structure. In the time dimension, the mountain tea garden is in a cloud cover zone throughout the year, so that the optical remote sensing data are frequently detected in a lack mode, and meanwhile, microclimate change caused by the cold air leakage and cold pool effect at night in the mountain has quick and short dynamic properties. When the sampling capability (spatial resolution or time acquisition frequency) of the monitoring system is lower than the structural frequency or dynamic change frequency of the monitored object, physical aliasing or folding phenomenon can occur. In this case, high frequency risk signals in terrace micro-terrains (such as diseases occurring in the part of the terrace wall or momentary cold air frost) can be erroneously folded into low frequency false fluctuations or background drift. The existing multi-source data fusion and blockchain tracing technology does not fully consider the observation distortion caused by sampling theorem failure. The system often records and chains the error signal after aliasing directly as a physical quantity that actually occurs. This approach not only results in misplacement of the pre-alarm locations in space-time, but also causes the blockchain system to solidify the responsibility for errors (e.g., false anomalies due to observation folds are due to mismanagement by tea farmers) into evidence that cannot be altered, thereby destroying the fairness and credibility of the traceability system. Disclosure of Invention The invention provides a full-flow digital early warning and blockchain responsibility tracing method for tea, which solves the technical problems in the background technology. The invention provides a full-flow digital early warning and blockchain responsibility tracing method for tea, which comprises the following steps: Constructing a terrace planting micro-unit distribution map, and analyzing terrace landform repetition intervals of each terrace planting micro-unit along the slope direction; acquiring multi-source heterogeneous sensing data aiming at the terraced field planting micro unit, setting a preset time window, calculating and determining comprehensive spatial resolution and comprehensive time acquisition interval in the preset time window, and acquiring continuous aging of a risk process of the terraced field planting micro unit; calculating to obtain a space-time sampling aliasing index based on the ratio relation between the comprehensive spatial resolution and the terrace landform repetition interval and the ratio relation between the comprehensive time acquisition interval and the continuous aging of the risk process; Generating a