CN116090834B - Forestry management method and device based on Flink platform - Google Patents
Forestry management method and device based on Flink platform Download PDFInfo
- Publication number
- CN116090834B CN116090834B CN202310206544.2A CN202310206544A CN116090834B CN 116090834 B CN116090834 B CN 116090834B CN 202310206544 A CN202310206544 A CN 202310206544A CN 116090834 B CN116090834 B CN 116090834B
- Authority
- CN
- China
- Prior art keywords
- monitoring
- risk type
- sub
- subarea
- region
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000007726 management method Methods 0.000 title claims abstract description 19
- 238000012544 monitoring process Methods 0.000 claims abstract description 274
- 239000013598 vector Substances 0.000 claims abstract description 57
- 238000003066 decision tree Methods 0.000 claims abstract description 54
- 238000000034 method Methods 0.000 claims abstract description 20
- 230000007613 environmental effect Effects 0.000 claims description 126
- 238000010276 construction Methods 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000011161 development Methods 0.000 claims description 3
- 238000012217 deletion Methods 0.000 claims 2
- 230000037430 deletion Effects 0.000 claims 2
- 230000002159 abnormal effect Effects 0.000 description 20
- CURLTUGMZLYLDI-UHFFFAOYSA-N Carbon dioxide Chemical compound O=C=O CURLTUGMZLYLDI-UHFFFAOYSA-N 0.000 description 18
- 229910002092 carbon dioxide Inorganic materials 0.000 description 9
- 239000001569 carbon dioxide Substances 0.000 description 9
- 230000005856 abnormality Effects 0.000 description 4
- 239000002245 particle Substances 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 241001465754 Metazoa Species 0.000 description 2
- 230000007812 deficiency Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 239000000779 smoke Substances 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/02—Agriculture; Fishing; Forestry; Mining
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Quality & Reliability (AREA)
- Animal Husbandry (AREA)
- Operations Research (AREA)
- Educational Administration (AREA)
- Game Theory and Decision Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Agronomy & Crop Science (AREA)
- Development Economics (AREA)
- Marine Sciences & Fisheries (AREA)
- Mining & Mineral Resources (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Alarm Systems (AREA)
Abstract
The invention relates to a forestry management method and device based on a Flink platform, wherein the method comprises the following steps: dividing a monitored target forest area into a plurality of monitoring subareas; extracting historical environment feature vectors in each monitoring subarea; selecting at least two monitoring subareas for multiple times to form a plurality of monitoring mother areas; based on the historical environment feature vectors corresponding to the monitoring subregions in the monitoring parent region, constructing a parent decision tree corresponding to the monitoring parent region so as to judge the risk type corresponding to the monitoring parent region; and determining the current risk type of the reference monitoring subarea based on the second risk type judgment results corresponding to all the monitoring mother areas comprising the reference monitoring subarea and the first risk type judgment results corresponding to the reference monitoring subarea.
Description
Technical Field
The invention relates to the field of forestry data processing, in particular to a forestry management method and device based on a Flink platform.
Background
At present, extreme weather often causes the occurrence of big fires or geological disasters in a forest system, and in order to effectively discover or predict the occurrence of the disasters, a large number of sensors and cameras are often arranged at various positions in the forest. However, a forest is a large-scale complex ecological and environmental system, and the damage of a sensor or a camera may be caused due to complex internal conditions, such as activities of some animals, excessive humidity of the environment, and the like, which directly results in insufficient monitoring or failure of a part of areas in the forest. If the sensor and the camera are replaced by entering the forest frequently, unnecessary interference to the ecological system in the forest is caused. Thus, the method is applicable to a variety of applications. How to analyze forestry data under the condition that a sensor or a camera is damaged, so that the occurrence of abnormal conditions in a forest is predicted and monitored, and the method is a troublesome problem.
Disclosure of Invention
Based on the above, it is necessary to provide a forest management method and device based on a link platform for solving the problem of insufficient monitoring in the case of damage of sensors or cameras in a forest.
A forestry management method based on a Flink platform comprises the following steps:
dividing a monitored target forest area into a plurality of monitoring subareas;
extracting historical environment feature vectors in all the monitoring subareas, and constructing sub-decision trees corresponding to all the monitoring subareas based on the historical environment feature vectors corresponding to the monitoring subareas so as to judge risk types corresponding to all the monitoring subareas;
selecting at least two monitoring subareas for multiple times to form a plurality of monitoring mother areas;
based on the historical environment feature vectors corresponding to the monitoring subregions in the monitoring parent region, constructing a parent decision tree corresponding to the monitoring parent region so as to judge the risk type corresponding to the monitoring parent region;
setting one of the monitoring subareas as a reference monitoring subarea, acquiring a current environment feature vector of the reference monitoring subarea, and inputting the current environment feature vector into a corresponding sub-decision tree in real time to acquire a first risk type judgment result;
acquiring current environment feature vectors of all monitoring subareas and inputting the current environment feature vectors into corresponding parent decision trees in real time so as to acquire second risk type judgment results corresponding to all monitoring parent areas;
and determining the current risk type of the reference monitoring subarea based on the second risk type judgment results corresponding to all the monitoring mother areas comprising the reference monitoring subarea and the first risk type judgment results corresponding to the reference monitoring subarea.
The invention provides a method for determining the boundary of a monitoring subarea by using a plurality of environment parameter detectors arranged in a target forest area, wherein each environment parameter detector is provided with a plurality of environment parameter detectors used for acquiring a plurality of environment characteristic parameters of specific types, comparing the environment characteristic parameters acquired by the environment parameter detectors of the same type under normal conditions to acquire environment characteristic parameter difference values, setting environment characteristic parameter difference value threshold values, and selecting the environment parameter detectors corresponding to environment characteristic parameter difference values lower than the environment characteristic parameter difference value threshold values to determine the boundary of the monitoring subarea.
According to the invention, at least part of the monitoring mother area is continuous in space, at least part of the monitoring mother area is distributed at intervals in the monitoring subareas, and the corresponding environmental characteristic parameter difference value is lower than the environmental characteristic parameter difference value threshold value.
The environmental parameter detector is one of a camera, a humidity sensor, a temperature sensor, a granularity sensor, a wind speed sensor, a temperature and humidity sensor and a carbon dioxide sensor.
The invention is based on the fly operation.
The invention compares the second risk type judgment results corresponding to all the monitoring parent areas comprising the reference monitoring subarea to predict the risk development trend in the reference monitoring subarea.
The invention carries out replacement sampling on the historical environmental characteristic vector corresponding to the monitoring subarea so as to obtain a plurality of sub-decision trees, constructs the current environmental characteristic vector of the reference monitoring subarea according to the environmental characteristic parameters obtained by the environmental characteristic detector which is not damaged currently in the reference monitoring subarea, and selects the sub-decision tree corresponding to the reference monitoring subarea based on the type of the environmental characteristic parameters in the current environmental characteristic vector of the reference monitoring subarea so as to obtain a first risk type judgment result.
The method comprises the steps of setting a parameter missing number threshold, counting the number of environmental characteristic parameters missing from a reference monitoring sub-region, and taking a first risk type judgment result as the current risk type of the reference monitoring sub-region if the number of missing environmental characteristic parameters is lower than the parameter missing number threshold.
If the missing quantity of the environmental characteristic parameters is not lower than the parameter missing quantity threshold value, counting all the second risk type judgment results to determine the current risk type of the reference monitoring subarea.
The invention sets the proportion threshold value b 3 If the missing quantity of the environmental characteristic parameters is not lower than the parameter missing quantity threshold value, the second risk type judgment result is that the normal quantity is b 1 The number of anomalies is b 2 If |b 1 -b 2 |/b 1 ＞b 3 Then with max { b } 1 ，b 2 Determining the current risk type of the reference monitor subarea.
The invention is that |b 1 -b 2 |/b 1 ≤b 3 The number of monitoring parent areas is increased until |b 1 -b 2 |/b 1 ＞b 3 。
A forestry management device, comprising:
the region dividing module is used for dividing the target forest region into a plurality of monitoring subareas;
the region combination module is used for combining the monitoring subareas into a monitoring mother region;
the sub-decision tree construction module is used for constructing sub-decision trees corresponding to all the monitoring subareas based on the historical environment feature vectors corresponding to the monitoring subareas;
the parent decision tree construction module is used for constructing a parent decision tree corresponding to the monitoring parent region based on the historical environment feature vectors corresponding to each monitoring sub-region in the monitoring parent region;
the calculation module inputs the current environment feature vector of the reference monitoring subarea into the corresponding sub-decision tree in real time to obtain a first risk type judgment result, and inputs the current environment feature vector of each monitoring subarea into the corresponding parent decision tree in real time to obtain a second risk type judgment result;
the comparison module is used for determining the current risk type of the reference monitoring subarea based on the second risk type judgment results corresponding to all monitoring mother areas comprising the reference monitoring subarea and the first risk type judgment results corresponding to the reference monitoring subarea.
A storage medium storing a program that when run performs a method of managing forestry based on a link platform.
The beneficial effects of the invention are as follows:
when the environmental characteristic parameters in the reference monitoring subarea are missing, the risk type judgment results of all monitoring master areas comprising the reference monitoring subarea are synthesized to reversely deduce the risk type of the reference monitoring subarea, and the possibility of abnormality of the reference monitoring subarea is estimated from probability, so that the risk type of the reference monitoring subarea is obtained in real time under the condition that the environment parameter detector in the reference monitoring subarea is not replaced, and meanwhile, the influence of human activities on a forest system is reduced.
Drawings
Figure 1 is a flow chart of a method for managing forestry based on a link platform in an embodiment of the present invention.
Detailed Description
In order that the above objects, features and advantages of the invention will be readily understood, a more particular description of the invention will be rendered by reference to the appended drawings. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. The present invention may be embodied in many other forms than described herein and similarly modified by those skilled in the art without departing from the spirit of the invention, whereby the invention is not limited to the specific embodiments disclosed below.
Furthermore, the terms "first," "second," and the like, are used for descriptive purposes only and are not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include at least one such feature. In the description of the present invention, the meaning of "plurality" means at least two, for example, two, three, etc., unless specifically defined otherwise.
Examples:
the target forest area is provided with a plurality of environment parameter detectors, and different types of environment parameter detectors are used for acquiring a plurality of specific types of environment characteristic parameters. For example, if the environmental parameter detector is a camera, the environmental characteristic parameter obtained by the environmental parameter detector is the number of fallen leaves and/or the number of trees and/or the number of animals, if the environmental parameter detector is a humidity sensor, the environmental characteristic parameter obtained by the environmental parameter detector is humidity, if the environmental parameter detector is a temperature sensor, the environmental characteristic parameter obtained by the environmental parameter detector is temperature, if the environmental parameter detector is a granularity sensor, the environmental characteristic parameter obtained by the environmental parameter detector is granularity in air, if the environmental parameter detector is a wind speed sensor, the environmental characteristic parameter obtained by the environmental parameter detector is wind speed, if the environmental parameter detector is a wind direction sensor, the environmental characteristic parameter obtained by the environmental parameter detector is wind direction, if the environmental parameter detector is a temperature and humidity sensor, and the environmental characteristic parameter obtained by the environmental parameter detector is carbon dioxide concentration. The wind direction sensor and the wind speed sensor are positioned at the same position to ensure that the measured flow speed and the measured flow direction of the wind correspond. The particle size sensor and the carbon dioxide sensor are generally arranged at a position close to the ground, and if a fire disaster occurs, smoke particles and carbon dioxide sink to the ground under the action of gravity, so that the smoke particles and the carbon dioxide are more easily detected by the particle size sensor and the carbon dioxide sensor. There are a plurality of each environmental parameter sensor to be as spread as possible throughout the forest.
Referring to fig. 1, based on this, this embodiment provides a forestry management method based on a link platform, including the following steps:
step S1: dividing a monitored target forest area into a plurality of monitoring subareas, wherein the space of the monitoring subareas is continuous, all the monitoring subareas are spliced to form the target forest area, and the dividing standard of the monitoring subareas can be set at will or can be based on content, for example, the whole lake is divided into one monitoring subarea, or the main distribution range of a certain type of tree is used as one monitoring subarea;
because the different monitoring subareas are positioned in the same target forest area, certain relevance exists among the environmental characteristic parameters of the different monitoring subareas, whether the different monitoring subareas are adjacent or spaced is irrelevant, and in order to better reflect the relevance in the subsequent process, the fact that the difference value among the environmental characteristic parameters in each part in the monitoring subareas cannot be too large is required to be ensured so as to eliminate the uncertainty of the relation among the environmental characteristic parameters in different positions in the monitoring subareas as far as possible;
based on this, the present step further includes step S11: comparing the environmental characteristic parameters obtained by the same type of environmental parameter detectors under normal conditions to obtain environmental characteristic parameter difference values, setting environmental characteristic parameter difference value threshold values, and selecting the environmental parameter detectors with the environmental characteristic parameter difference values lower than the environmental characteristic parameter difference value threshold values to determine the boundary of the monitoring subarea;
for example, three humidity sensors in a substantially straight line, the humidity values obtained in sequence are a respectively 1 、a 2 、a 3 The environmental characteristic parameter difference threshold A, if |a 1 -a 2 I < A, and I a 1 -a 3 I > A, then the value of the passing humidity may be a 2 The straight line or the curve of the humidity sensor is a boundary line and is used as the boundary line of the monitoring subarea;
step S2: extracting historical environment feature vectors in all the monitoring subareas, and constructing sub-decision trees corresponding to all the monitoring subareas based on the historical environment feature vectors corresponding to the monitoring subareas so as to judge risk types corresponding to all the monitoring subareas;
the historical environment characteristic vector is composed of environment characteristic parameters such as a historical temperature value, a historical humidity value, a historical granularity, a historical carbon dioxide concentration and the like of the monitoring subarea at the same moment, for example, the historical humidity value is too low, which indicates that the monitoring subarea is dry at the moment and has the risk of fire, or the historical carbon dioxide concentration and the historical temperature value are too high, which indicates that the monitoring subarea possibly has fire at the moment, and the information such as the fire occurrence degree and the like can be obtained according to the values of the historical carbon dioxide concentration and the historical temperature value; the sub decision tree generated by a plurality of environmental characteristic parameters in the historical environmental characteristic vector can judge the risk type corresponding to the monitoring subarea through the environmental characteristic parameters, the risk type can be in a normal state, a state that fire is easy to occur, a state that fire is occurring and the like, and the risk type that fire is easy to occur, a state that fire is occurring and the like can be reduced to an abnormal state;
it can be understood that, under the condition that the historical environment feature vector is complete, the accuracy of the risk type obtained by judging by the monitoring sub-region by utilizing the sub-decision tree corresponding to the monitoring sub-region is theoretically highest, so that the construction process of the sub-decision tree needs to use the historical environment feature vector which is as complete as possible;
step S3: selecting at least two monitoring subareas for multiple times to form a plurality of monitoring mother areas;
at least two monitoring subareas are selected at a time to form a monitoring master area, different monitoring master areas can be formed by selecting different monitoring subareas or changing the number of the selected monitoring subareas, and a plurality of different monitoring master areas can be correspondingly obtained by repeating the selecting process; the risk type in the monitoring parent region is also related to the risk type of each monitoring subarea in the monitoring parent region, and at the same time, any monitoring subarea is required to be ensured to belong to at least two different monitoring parent regions at the same time;
the correlation of the environmental characteristic parameters between different monitoring subareas is strong and weak, and the correlation difference between the environmental characteristic parameters of the corresponding monitoring master area and the environmental characteristic parameters of each monitoring subarea in the corresponding monitoring master area is also generated, so that the correlation of the environmental characteristic parameters between any two monitoring subareas in the same monitoring master area needs to be effectively increased in order to ensure that the environmental characteristic parameters of the monitoring master area and the environmental characteristic parameters of each monitoring subarea in the monitoring master area are both strong and strong;
based on this, in this embodiment, at least part of the monitoring parent regions are spatially continuous, that is, part of the two adjacent or sequentially adjacent monitoring sub-regions form the monitoring parent region, and the environmental characteristic parameter difference value is higher than the environmental characteristic parameter difference value threshold value in the part of the monitoring parent region due to the fact that the distance between the monitoring sub-regions is very close to each other, but there is still a considerable correlation, that is, the environmental characteristic parameter of the adjacent monitoring sub-regions is affected soon under the condition that the environmental characteristic parameter of one of the monitoring sub-regions is abnormally changed;
on the other hand, in at least part of the monitoring parent areas, the monitoring subareas are distributed at intervals and the corresponding environmental characteristic parameter difference value is lower than the environmental characteristic parameter difference value, and for the monitoring subareas in the part of the monitoring parent areas, although the positions of the monitoring subareas are not adjacent, the relevance between the monitoring subareas is weakened, and the monitoring subareas are all lakes, or the main plants are of the same kind, or the vegetation density is similar, so that the environmental characteristic parameters of the monitoring subareas are similar under normal conditions, and if the environmental characteristic parameters of the two monitoring subareas deviate greatly, the abnormal condition of at least one monitoring subarea is indicated;
step S4: after the construction of the monitoring master area is completed, constructing a master decision tree corresponding to the monitoring master area based on the historical environment feature vectors corresponding to each monitoring sub-area in the monitoring master area so as to judge the risk type corresponding to the monitoring master area;
in the construction process of the parent decision tree, all the environmental characteristic parameters in the historical environmental characteristic vectors corresponding to each monitoring subarea are not used, but partial environmental characteristic parameters are respectively extracted from the historical environmental characteristic vectors of each monitoring subarea to form a new historical environmental characteristic vector, and the parent decision tree is constructed based on the new historical environmental characteristic vector;
the risk type judging result of the master decision tree is related to the judging result of the sub decision tree corresponding to each monitoring subarea in the monitoring master area, and the judging result of the sub decision tree corresponding to each monitoring subarea in the monitoring master area is finally reflected to the risk type judging result of the master decision tree;
step S5: setting one of the monitoring subareas as a reference monitoring subarea, acquiring a current environment feature vector of the reference monitoring subarea, and inputting the current environment feature vector into a corresponding sub-decision tree in real time to acquire a first risk type judgment result;
the current environmental feature vector of the reference monitoring subarea theoretically contains the environmental feature parameters monitored by all the environmental parameter detectors in the reference monitoring subarea, but when part of the environmental parameter detectors in the reference monitoring subarea are damaged or fail, the acquisition of part of the environmental feature parameters is failed, so that the current environmental feature vector of the reference monitoring subarea is quite possibly incomplete, the actual inaccurate first risk type judgment result is caused, and the accuracy of the first risk type judgment result is lower as the missing environmental feature parameters are more;
step S6: acquiring current environment feature vectors of all monitoring subareas and inputting the current environment feature vectors into corresponding parent decision trees in real time so as to acquire second risk type judgment results corresponding to all monitoring parent areas;
if the monitoring parent region is continuous in space, the second risk type judgment result is abnormal as long as the first risk type judgment results corresponding to the plurality of monitoring subregions in the monitoring parent region are abnormal, otherwise, if the second risk type judgment results are abnormal, the first risk type judgment results corresponding to at least part of the monitoring subregions in the monitoring parent region are also abnormal;
if the monitoring subareas of the monitoring mother area are distributed at intervals, the second risk type judgment result is abnormal as long as the first risk type judgment results corresponding to the plurality of monitoring subareas in the monitoring mother area are abnormal, otherwise, if the second risk type judgment result is abnormal, the first risk type judgment result corresponding to at least part of the monitoring subareas in the monitoring mother area is also abnormal; additionally, if the second risk type judgment result in the monitoring parent region is abnormal, the abnormal judgment result may be caused by a large difference of environmental characteristic parameters among different monitoring subregions, and if the first risk type judgment results corresponding to the monitoring subregions are normal, a large probability that a plurality of first risk type judgment results are wrong is indicated;
based on the analysis inference, the identification can be performed under the condition that the first risk type judgment result is misjudged to be normal by counting the number or the proportion of the second risk type judgment result to be normal or abnormal;
step S7: determining the current risk type of the reference monitoring subarea based on the second risk type judgment results corresponding to all monitoring mother areas comprising the reference monitoring subarea and the first risk type judgment results corresponding to the reference monitoring subarea;
the loss of part of environmental characteristic parameters in the current environmental characteristic vector of the reference monitoring subarea also directly affects the judgment accuracy of the second risk type judgment result corresponding to the monitoring mother area containing the reference monitoring subarea;
however, since the monitoring parent region further includes other monitoring subregions having strong correlations with the reference monitoring subregion, part of the environmental characteristic parameters in the part of the monitoring subregions having strong correlations can partially fill the influence of the environmental characteristic parameter deficiency in the reference monitoring subregion, so that the accuracy of the second risk type judgment result is not reduced too low, or as long as the number of the monitoring subregions in the monitoring parent region is sufficiently large, the probability that all or most of the monitoring subregions have the environmental characteristic parameter deficiency is relatively low, so that the accuracy of the second risk type judgment result is not excessively reduced;
based on the analysis process of the second risk type judgment result in step S6, the judgment process of the current risk type of the reference monitoring sub-area in step S7 has the following cases:
1. when part of environmental characteristic parameters in the current environmental characteristic vector of the reference monitoring subarea are missing, but the number of the missing parameters is smaller, the first risk type judgment result corresponding to the reference monitoring subarea still has higher reliability, and based on the deduction, the current risk type in the reference monitoring subarea is determined to be in reference to the first risk type judgment result corresponding to the reference monitoring subarea;
specifically, a parameter missing number threshold is set, the quantity of environmental characteristic parameters missing from the reference monitoring sub-region is counted, and if the quantity of missing environmental characteristic parameters is lower than the parameter missing number threshold, the first risk type judgment result is used as the current risk type of the reference monitoring sub-region.
The judging accuracy of the sub-decision tree is determined by the node condition types, namely the number of the environmental characteristic parameter types used for judging, and the more the number of the environmental characteristic parameter types used in the judging process of the sub-decision tree is, the higher the accuracy of the first risk type judging result is; the historical environmental feature vectors corresponding to the monitoring subareas are replaced and sampled to obtain a plurality of sub-decision trees, so that at least part of the sub-decision trees are decided according to not all environmental feature parameters, the current environmental feature vector of the reference monitoring subareas is constructed according to the environmental feature parameters obtained by the environmental feature parameter detectors which are not damaged currently in the reference monitoring subareas, the sub-decision trees corresponding to the reference monitoring subareas are selected based on the types of the environmental feature parameters in the current environmental feature vector of the reference monitoring subareas, and the sub-decision trees still have higher accuracy in order to obtain the first risk type judgment result;
for example, the total number of the complete environmental characteristic parameter types is four, I, II, III and IV respectively, and substitution sampling is adopted, so that the judging process of some sub-decision trees depends on I, II and III instead of IV, if the current IV is just missing, the sub-decision trees which depend on I, II and III are selected, the missing of IV can not cause the accuracy of the first risk type judging result of the sub-decision tree to be reduced, and of course, the condition is only suitable for the situation that the number of the missing environmental characteristic parameter types is less, for example, the number of the missing environmental characteristic parameter types only accounts for less than 30% of the number of the total environmental characteristic parameter types, otherwise, the number of the environmental characteristic parameter types which can be used for judging by the sub-decision trees is too small, so that the accuracy of the first risk type judging result cannot meet the basic requirement;
2. when a large number of environmental characteristic parameters in the current environmental characteristic vector of the reference monitor sub-region are missing, the exemplary missing proportion is more than 30%, then the correctness of the first risk type judgment result corresponding to the reference monitor sub-region is doubtful, at the moment, if the second risk type judgment result corresponding to most of the monitoring parent regions comprising the reference monitor sub-region is abnormal, the exemplary abnormal proportion is more than 55%, then the current risk type of the reference monitor sub-region is regarded as abnormal, if the second risk type judgment result corresponding to most of the monitoring parent regions comprising the reference monitor sub-region is normal, the exemplary abnormal proportion is more than 55%, then the current risk type of the reference monitor sub-region is regarded as normal;
3. when a large number of environmental feature parameters in the current environmental feature vector of the reference monitor sub-region are missing, the exemplary missing proportion reaches more than 30%, then the correctness of the first risk type judgment result corresponding to the reference monitor sub-region is doubtful, at this time, if the second risk type judgment result corresponding to the monitor parent region including the reference monitor sub-region is similar in number to the abnormality, and the exemplary difference between the proportion of the second risk type judgment result to the abnormality is smaller than 10%, then it is difficult to judge whether the second risk type judgment result is caused by the reference monitor sub-region or not, at this time, step S3 may be performed again, and the number of the monitor parent regions including the reference monitor sub-region may be further increased until the number of the second risk type judgment result to the abnormality and the normal number of the first risk type judgment result is pulled apart by a sufficient gap, for example, the number gap is not smaller than 10%, at this time, the risk type of the reference monitor sub-region is judged based on the statistical result of the second risk type judgment result.
And repeating the process, and continuously selecting different monitoring subareas as reference monitoring subareas, so that risk type judgment results of different monitoring subareas with higher accuracy can be obtained in real time.
If the reference monitoring subarea is abnormal, the second risk type judgment result corresponding to the monitoring mother area containing the reference monitoring subarea is also abnormal, and the probability is inversely related to the number of the monitoring subareas existing in the monitoring mother area and is positively related to the area or the space ratio of the reference monitoring subarea to the monitoring mother area. Meanwhile, the monitoring mother region which comprises the reference monitoring sub region and is continuous in space can be regarded as a space formed by extending the reference monitoring sub region to a certain direction or a plurality of directions, and the extending directions of the reference monitoring sub regions corresponding to different monitoring mother regions can be the same or different. Based on the above reasoning, the following results can be obtained: if the number of the second risk type judgment results corresponding to the monitoring mother region formed by extending the reference monitoring sub-region in the first direction is larger, and the number of the second risk type judgment results corresponding to the monitoring mother region formed by extending the reference monitoring sub-region in the second direction is smaller, the risk development trend in the reference monitoring sub-region is indicated to be more probable along the first direction.
In this embodiment, the calculation amount of the forest management method based on the link platform is large, and the environmental characteristic parameter is unbounded data, so that the analysis method is based on link operation.
The embodiment also provides a forestry management device based on the above-mentioned Flink platform-based forestry management method, which comprises:
the region dividing module is used for dividing the target forest region into a plurality of monitoring subareas;
the region combination module is used for combining the monitoring subareas into a monitoring mother region;
the sub-decision tree construction module is used for constructing sub-decision trees corresponding to all the monitoring subareas based on the historical environment feature vectors corresponding to the monitoring subareas;
the parent decision tree construction module is used for constructing a parent decision tree corresponding to the monitoring parent region based on the historical environment feature vectors corresponding to each monitoring sub-region in the monitoring parent region;
the calculation module inputs the current environment feature vector of the reference monitoring subarea into the corresponding sub-decision tree in real time to obtain a first risk type judgment result, and inputs the current environment feature vector of each monitoring subarea into the corresponding parent decision tree in real time to obtain a second risk type judgment result;
the comparison module is used for determining the current risk type of the reference monitoring subarea based on the second risk type judgment results corresponding to all monitoring mother areas comprising the reference monitoring subarea and the first risk type judgment results corresponding to the reference monitoring subarea.
The embodiment also provides a storage medium which stores a program and executes the forestry management method based on the Flink platform when the program runs.
The technical features of the above-described embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above-described embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The above examples illustrate only a few embodiments of the invention, which are described in detail and are not to be construed as limiting the scope of the invention. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the invention, which are all within the scope of the invention. Accordingly, the scope of protection of the present invention is to be determined by the appended claims.
Claims (10)
1. A forestry management method based on a Flink platform is characterized by comprising the following steps:
dividing a monitored target forest area into a plurality of monitoring subareas;
extracting historical environment feature vectors in all the monitoring subareas, and constructing sub-decision trees corresponding to all the monitoring subareas based on the historical environment feature vectors corresponding to the monitoring subareas so as to judge risk types corresponding to all the monitoring subareas;
selecting at least two monitoring subareas for multiple times to form a plurality of monitoring mother areas;
based on the historical environment feature vectors corresponding to the monitoring subregions in the monitoring parent region, constructing a parent decision tree corresponding to the monitoring parent region so as to judge the risk type corresponding to the monitoring parent region;
setting one of the monitoring subareas as a reference monitoring subarea, acquiring a current environment feature vector of the reference monitoring subarea, and inputting the current environment feature vector into a corresponding sub-decision tree in real time to acquire a first risk type judgment result;
acquiring current environment feature vectors of all monitoring subareas and inputting the current environment feature vectors into corresponding parent decision trees in real time so as to acquire second risk type judgment results corresponding to all monitoring parent areas;
and determining the current risk type of the reference monitoring subarea based on the second risk type judgment results corresponding to all the monitoring mother areas comprising the reference monitoring subarea and the first risk type judgment results corresponding to the reference monitoring subarea.
2. The method for managing forest based on the link platform according to claim 1, wherein a plurality of environmental parameter detectors are provided in the target forest area, each of the plurality of environmental parameter detectors is used for acquiring a plurality of environmental characteristic parameters of a specific type, the environmental characteristic parameters obtained under normal conditions by the same type of environmental parameter detectors are compared to obtain environmental characteristic parameter differences, environmental characteristic parameter difference thresholds are set, and environmental parameter detectors corresponding to the environmental characteristic parameter difference values lower than the environmental characteristic parameter difference thresholds are selected to determine the boundary of the monitoring subarea.
3. A link platform based forestry management method according to claim 2, wherein at least a portion of the monitored parent region is spatially continuous, at least a portion of the monitored parent region is within the monitored sub-region, the monitored sub-regions are spaced apart and the corresponding environmental characteristic parameter differences are below an environmental characteristic parameter difference threshold.
4. A link platform-based forestry management method according to claim 1, wherein the second risk type judgment results corresponding to all the monitoring parent regions including the reference monitoring sub-region are compared with each other to predict a risk development trend in the reference monitoring sub-region.
5. The link platform-based forestry management method of claim 1, wherein historical environmental feature vectors corresponding to the monitoring sub-regions are sampled back to obtain a plurality of sub-decision trees, and the sub-decision trees corresponding to the reference monitoring sub-regions are selected based on the types of environmental feature parameters in the current environmental feature vectors of the reference monitoring sub-regions to obtain a first risk type judgment result.
6. The method for managing forestry based on a link platform according to claim 1, wherein a parameter deletion number threshold is set, the number of environmental characteristic parameters deleted from the reference monitor sub-area is counted, and if the number of environmental characteristic parameters deleted is lower than the parameter deletion number threshold, the first risk type judgment result is used as the current risk type of the reference monitor sub-area.
7. A method for managing forestry based on a Flink platform as defined in claim 6, wherein if the number of missing environmental characteristic parameters is not lower than a parameter missing number threshold, all second risk type judgment results are counted to determine the current risk type of the reference monitoring sub-region.
8. A link platform based forestry management method according to claim 7, wherein a scaling threshold b is set 3 If the missing quantity of the environmental characteristic parameters is not lower than the parameter missing quantity threshold value, the second risk type judgment result is that the normal quantity is b 1 The number of anomalies is b 2 If |b 1 -b 2 |/b 1 ＞b 3 Then with max { b } 1 ，b 2 Determining the current risk type of the reference monitor subarea.
9. A method of link platform based forestry management as recited in claim 8, wherein if ib 1 -b 2 |/b 1 ≤b 3 The number of monitoring parent areas is increased until |b 1 -b 2 |/b 1 ＞b 3 。
10. A forestry management device, comprising:
the region dividing module is used for dividing the target forest region into a plurality of monitoring subareas;
the region combination module is used for combining the monitoring subareas into a monitoring mother region;
the sub-decision tree construction module is used for constructing sub-decision trees corresponding to all the monitoring subareas based on the historical environment feature vectors corresponding to the monitoring subareas;
the parent decision tree construction module is used for constructing a parent decision tree corresponding to the monitoring parent region based on the historical environment feature vectors corresponding to each monitoring sub-region in the monitoring parent region;
the calculation module inputs the current environment feature vector of the reference monitoring subarea into the corresponding sub-decision tree in real time to obtain a first risk type judgment result, and inputs the current environment feature vector of each monitoring subarea into the corresponding parent decision tree in real time to obtain a second risk type judgment result;
the comparison module is used for determining the current risk type of the reference monitoring subarea based on the second risk type judgment results corresponding to all monitoring mother areas comprising the reference monitoring subarea and the first risk type judgment results corresponding to the reference monitoring subarea.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310206544.2A CN116090834B (en) | 2023-03-07 | 2023-03-07 | Forestry management method and device based on Flink platform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310206544.2A CN116090834B (en) | 2023-03-07 | 2023-03-07 | Forestry management method and device based on Flink platform |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116090834A CN116090834A (en) | 2023-05-09 |
| CN116090834B true CN116090834B (en) | 2023-06-13 |
Family
ID=86199342
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310206544.2A Active CN116090834B (en) | 2023-03-07 | 2023-03-07 | Forestry management method and device based on Flink platform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116090834B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2631866A1 (en) * | 2012-02-23 | 2013-08-28 | UPM-Kymmene Corporation | A method for determining the risk of forest destruction, and a method for forest management |
| WO2015001558A1 (en) * | 2013-07-01 | 2015-01-08 | Salespredict Sw Ltd. | System and method for predicting sales |
| US9760978B1 (en) * | 2016-05-09 | 2017-09-12 | Adobe Systems Incorporated | Missing region prediction |
| CN109636069A (en) * | 2019-01-29 | 2019-04-16 | 平安科技（深圳）有限公司 | Data processing method, device, equipment and storage medium |
| AU2020100709A4 (en) * | 2020-05-05 | 2020-06-11 | Bao, Yuhang Mr | A method of prediction model based on random forest algorithm |
| CN111292020A (en) * | 2020-03-13 | 2020-06-16 | 贵州电网有限责任公司 | Power grid real-time operation risk assessment method and system based on random forest |
| CN111930049A (en) * | 2020-08-21 | 2020-11-13 | 广州立信电子科技有限公司 | Forestry forest fire prevention safety intelligent monitoring management system based on thing networking |
| WO2022104736A1 (en) * | 2020-11-20 | 2022-05-27 | 周连惠 | Method and system for managing forest resources |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7328218B2 (en) * | 2005-03-22 | 2008-02-05 | Salford Systems | Constrained tree structure method and system |
| US10922701B2 (en) * | 2016-07-28 | 2021-02-16 | Mastercard International Incorporated | Systems and methods for characterizing geographic regions |
-
2023
- 2023-03-07 CN CN202310206544.2A patent/CN116090834B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2631866A1 (en) * | 2012-02-23 | 2013-08-28 | UPM-Kymmene Corporation | A method for determining the risk of forest destruction, and a method for forest management |
| WO2015001558A1 (en) * | 2013-07-01 | 2015-01-08 | Salespredict Sw Ltd. | System and method for predicting sales |
| US9760978B1 (en) * | 2016-05-09 | 2017-09-12 | Adobe Systems Incorporated | Missing region prediction |
| CN109636069A (en) * | 2019-01-29 | 2019-04-16 | 平安科技（深圳）有限公司 | Data processing method, device, equipment and storage medium |
| CN111292020A (en) * | 2020-03-13 | 2020-06-16 | 贵州电网有限责任公司 | Power grid real-time operation risk assessment method and system based on random forest |
| AU2020100709A4 (en) * | 2020-05-05 | 2020-06-11 | Bao, Yuhang Mr | A method of prediction model based on random forest algorithm |
| CN111930049A (en) * | 2020-08-21 | 2020-11-13 | 广州立信电子科技有限公司 | Forestry forest fire prevention safety intelligent monitoring management system based on thing networking |
| WO2022104736A1 (en) * | 2020-11-20 | 2022-05-27 | 周连惠 | Method and system for managing forest resources |
Non-Patent Citations (4)
| Title |
|---|
| 一种改进的决策树分类方法在土地利用信息提取中的应用;曾维军;侯明明;杨伟;;贵州大学学报(自然科学版)(第06期);第39-46页 * |
| 合肥市不同类型城市森林树种多样性比较;许克福;吴泽民;陈家龙;;东北林业大学学报(第03期);第26-30页 * |
| 基于决策树的地下工程透水事故发生风险定量评估;刘建坡 等;《金属矿山》;第217-224页 * |
| 基于随机森林的山洪灾害风险评估――以江西省为例;吴小君;方秀琴;任立良;吴陶樱;苗月鲜;;水土保持研究(第03期);第142-149页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116090834A (en) | 2023-05-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10373065B2 (en) | Generating database cluster health alerts using machine learning | |
| Özbayoğlu et al. | Estimation of the burned area in forest fires using computational intelligence techniques | |
| CN111898647B (en) | Clustering analysis-based low-voltage distribution equipment false alarm identification method | |
| CN116523508B (en) | Intelligent operation and maintenance management system based on AIoT asset | |
| GB2522926A (en) | Method of identifying anomalies | |
| CN115118581B (en) | Internet of things data all-link monitoring and intelligent guaranteeing system based on 5G | |
| CN115865649B (en) | Intelligent operation and maintenance management control method, system and storage medium | |
| CN108572308A (en) | fault diagnosis method and system | |
| CN113781732A (en) | Cable channel fire early warning method and device based on intelligent gateway | |
| CN111176872B (en) | Monitoring data processing method, system, device and storage medium for IT operation and maintenance | |
| Doostan et al. | Power distribution system equipment failure identification using machine learning algorithms | |
| CN117148045A (en) | Fault studying and judging management system for running state of power distribution network | |
| CN116090834B (en) | Forestry management method and device based on Flink platform | |
| US9654997B2 (en) | Method and system for evaluation of sensor observations | |
| Harrou et al. | Enhanced monitoring using PCA-based GLR fault detection and multiscale filtering | |
| Wang et al. | Forest fire detection system based on Fuzzy Kalman filter | |
| CN111306051B (en) | Probe type state monitoring and early warning method, device and system for oil transfer pump unit | |
| CN116504016B (en) | Thermal power plant safety monitoring and early warning method and system based on artificial intelligence | |
| CN110007171A (en) | The screening method and system of transformer online monitoring data false alarm | |
| CN108829147A (en) | Mine exhaust management method and system based on Internet of Things | |
| FR2878637A1 (en) | DELETING FALSE ALERTS AMONG ALERTS PRODUCED IN A MONITORED INFORMATION SYSTEM | |
| CN109558258B (en) | Method and device for positioning root fault of distributed system | |
| CN113489929A (en) | Method and device for monitoring supervision of garbage classification, computer equipment and storage medium | |
| US20150347213A1 (en) | Apparatus and method for system monitoring | |
| CN117439899B (en) | Communication machine room inspection method and system based on big data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information |
Inventor after: Lin Chen Inventor after: Huang Yu Inventor before: Lin Chen |