CN110677454B - Water pollution early warning system and method based on multi-agent network convergence algorithm - Google Patents

Abstract

The invention relates to the technical field of environmental monitoring, in particular to a water pollution early warning system and method based on a multi-agent network convergence control algorithm, comprising the following steps: (1): consider a system consisting of n intelligent agent nodes (monitoring points) (2): Select z water environment evaluation standards, and select the corresponding category standards according to the water functional categories; configure one or more nodes for the nodes (3): Using a distributed control protocol, when the state value of the water area monitored by the node exceeds the standard range, the node determines that the water environment is polluted, thereby triggering an early warning. The method can effectively improve the response speed of the system, reduce the energy consumption of equipment, improve the reliability of system operation, and reduce the manufacturing cost of the overall system, and has considerable market prospects in the field of water pollution environment monitoring and early warning.

Description

Water pollution early warning system and method based on multi-agent network convergence algorithm

technical field

The invention relates to the technical field of environmental monitoring, in particular to a water pollution early warning system and method based on a multi-agent network convergence control algorithm.

Background technique

With the continuous development of social economy and the continuous improvement of people's living standards, the problem of water pollution has also become very serious. A large amount of industrial, agricultural and domestic waste water is directly discharged into rivers, lakes and oceans without treatment, causing serious pollution to water quality, and at the same time causing great harm to ecological environment security, social stability, human health and national property security. damage. In the water pollution monitoring system, it can be divided into three levels: rapid early warning, deterministic detection and accurate authoritative detection. The rapid early warning link is the basis of the latter two levels of detection, which is important for striving for the best time for treatment and reducing major environmental and economic losses. Crucial role.

The traditional sensor network-based water pollution early warning system is basically realized through a centralized communication system, that is, each monitoring node is required to have the ability to communicate with the monitoring center node. The main defect of this type of system is that the workload of the monitoring central node station is too large, and once the central node strikes due to network failure, congestion, or even malicious network attacks, the entire water pollution early warning system will be paralyzed, and it has poor robustness. sex. At the same time, because water pollution monitoring and early warning is a long-term work, in order to keep the system working normally, the equipment must be replaced frequently (node energy exhaustion and failure), which not only has a large maintenance workload but also increases operating costs. To a certain extent, it limits its use as a water pollution detection and early warning technology.

Multi-Agent Networks refers to a network composed of a large number of individual agents with local perception, execution, and communication capabilities. In recent years, multi-agent networks have been widely used in multi-robot systems, smart grid scheduling, unmanned aerial vehicles, and wireless sensor networks due to their advantages of robustness, autonomy, and low cost. Convergence algorithm is a kind of important distributed algorithm in current multi-agent network, which means that the nodes in the network communicate and cooperate with their neighbor nodes without the need of a central node, and finally realize that the state values of all nodes in the system tend to be the same. status value.

The disclosure of the above background technology content is only used to assist the understanding of the inventive concept and technical solution of the present invention, and it does not necessarily belong to the prior art of this patent application. If there is no clear evidence that the above content has been disclosed on the filing date of this patent application The above background art should not be used to evaluate the novelty and inventive step of the present application.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a water pollution early warning method based on a multi-agent network convergence algorithm. By adopting this method, the response speed of the system can be effectively improved, the energy consumption of equipment can be reduced, the operation reliability of the system can be improved, and the manufacturing cost of the overall system can be reduced at the same time. , has considerable market prospects in the field of water pollution environmental monitoring and early warning.

In order to achieve the above objects, the technical solution adopted by the present invention includes the following three aspects.

In the first aspect, a water pollution early warning method based on a multi-agent network convergence algorithm, comprising the following steps:

Step (1): Consider a multi-agent network composed of n agent nodes (monitoring points), and construct the entire network topology into a connected graph according to the node communication edge relationship, that is, starting from any node in the network, through the relative A directed edge of a neighboring node can reach any other node in the network;

Step (2): According to the "Surface Water Environmental Quality Standard" (GB3838-2002), select z water environment evaluation standards, and select the corresponding category standards according to the water area function category; configure one or more sensors for the corresponding monitoring indicators for the nodes;

Step (3): Assuming that N _i (t) is the set of neighbor nodes of node i at time t, a distributed control protocol is adopted, and the design of the control protocol follows the following rules: node i organizes the information sent by neighbor node j collected at time t. , then when the state value of the water area monitored by the node exceeds the standard interval, the node determines that the water environment is polluted, thereby triggering an early warning, and the corresponding controller triggers intervention.

It is worth noting that the present invention only involves the information of at most one neighbor node exceeding the threshold at the same time in the system, that is, only one of the situations in which the state value of a node is higher than the upper limit of the threshold or lower than the lower limit of the threshold occurs in the system. Consider the simultaneous occurrence of both cases. However, the system and method of the present application have the following advantages: compared with the existing design method, only a single state information of adjacent nodes needs to be transmitted, which reduces the communication volume in the system and reduces the communication pressure of the system; The event, that is, the monitoring value triggers the early warning mechanism, which reduces the energy consumption of the system; when designing the convergence control protocol, that is, step 3 of this method, each node only needs to select the largest or smallest neighbor value as its own state value at the next moment, This results in a distributed event-driven consensus protocol. Compared with the traditional method, the design process of the control protocol is greatly simplified; the implementation of the threshold trigger mechanism of the present invention and the early warning algorithm based on this mechanism only needs to use the node and the neighbors in the network that can exchange information with the node. The overall design method of node information is based on distributed control architecture, which has the characteristics of low cost, strong scalability and high robustness, and is especially suitable for large-scale water source environment monitoring and early warning when a large number of distributed sensors are connected.

表示节点间的通讯边集；(i,j)∈E表示节点i可接收到来自节点j的信息。Further, in step (1), the multi-agent network composed of n agent nodes (monitoring points) can be denoted as G=(V,E), where V={1,2,...,n} represents a node set,

Represents the communication edge set between nodes; (i,j)∈E indicates that node i can receive information from node j.

Further, in step (2), let X=[x _i,k , k, Lng _i , Lat _i ], k∈Z, Z={1,2,...,z}, where x _{i, k} represents the kth monitoring index (ie, status value) of the ith node, Lng _i represents the longitude of the node i, and Lat _i represents the latitude of the node i.

Further, the method for selecting the corresponding category standard according to the water area function category described in step (2) includes the following steps:

Step (2.1): Determine the water function category c∈C to which the water body belongs, C={c1,c2,c3,c4,c5}={Class I water, Class II water, Class III water, Class IV water, Class V water };

其中

x _c,k分别表示c类水第k个监测指标的标准上限和标准下限，c∈C。Step (2.2): Select the corresponding standard of the corresponding functional category, and set up the monitoring standard value range

in

x _{c, k} represent the upper and lower standard limits of the k-th monitoring indicator of water in category c, respectively, c∈C.

Further, in step (3),

set of neighbor nodes

Distributed control protocol U={u ₁ ,u ₂ ,...,u _n };

则当节点监测的水域状态值超出标准区间时，预警系统启动，即当

时，或x_k,m(t)<x _k时，与之相应的控制器触发介入，具体控制器协议为

或

The design of the control protocol is based on the following rules: node i organizes the information sent by neighbor nodes j, j∈N _i collected at time t, so that

Then, when the water state value monitored by the node exceeds the standard range, the early warning system is activated, that is, when

, or when x _k,m (t)< x _k , the corresponding controller triggers intervention, and the specific controller protocol is

or

After the controller intervenes, the state value of each node in the system is updated.

Further, in step (3), after the controller intervenes, the update equation of the state value of each node in the system is as follows:

则节点不采取任何措施；而当满足

条件时，该节点判定水环境受到污染，从而触发预警，该时刻起，此后节点i将其状态值维持x_i,k(t)固定不变，并将此固定值传递给其相邻节点；网络中其余节点则按照公式(1)更新自身下一时刻的状态值，并同步节点信息X_i,k(t+1)＝[x_i,k(t+1),k,经度坐标i,纬度坐标i]，并在全网络系统传递污染坐标地址和污染的状态值，直到完成同步全部网络节点。Further, it also includes step (4): node i collects and obtains the monitoring state value x _i,k (t) of its own area at time t; when the state value

then the node does not take any action; and when the

When the conditions are met, the node determines that the water environment is polluted, thereby triggering an early warning. From this moment on, node i keeps its state value x _i,k (t) fixed, and transmits this fixed value to its adjacent nodes; The remaining nodes in the network update their state values at the next moment according to formula (1), and synchronize the node information Xi _,k (t+1)=[ _xi,k (t+1),k, longitude coordinate i, Latitude coordinate i], and transmit the pollution coordinate address and pollution status value in the whole network system until the synchronization of all network nodes is completed.

Further, the method for synchronous transmission of multi-agent network information described in step (4) includes the following steps:

Step (4.1): The node collects the data set x _i,k (t) at time t through the corresponding sensor, and obtains the latitude and longitude information through its own GPS positioning (or manually given according to the location of the node in advance);

Step (4.2): When the number monitored by node i is within the normal range, that is, it does not exceed the set threshold range, node i does not take any measures; if the state value measured by node i exceeds the threshold range, node i records An outlier at this moment, and this number is fixed for all subsequent times;

Step (4.3): After node i measures the abnormal value, it triggers the early warning mechanism, and starts to use the wireless communication protocol to send the abnormal state value to its adjacent nodes;

Step (4.4): The adjacent node j updates its own state value based on the maximum/minimum convergence algorithm, and at the same time transmits the updated state value to its adjacent nodes;

Step (4.5): Through the local information interaction, the outliers are continuously transmitted to the diffusion, until the outliers are synchronized to all nodes in the whole network to be consistent.

The second aspect, a water pollution early warning system based on a multi-agent network convergence algorithm, including:

-Monitoring module: it includes a multi-agent network composed of n agent nodes (monitoring points), and the entire network topology is constructed as a connected graph;

- Sensor module: It is configured at the node and can detect the corresponding indicators of the corresponding node. According to the "Surface Water Environmental Quality Standard" (GB3838-2002), z water environment evaluation standards are selected, and the corresponding category standards are selected according to the functional categories of water areas;

- Control module: it contains a distributed control protocol, the protocol is used to determine that the water environment is polluted when the state value of the water area monitored by the node exceeds the standard range, thereby triggering an early warning, and the corresponding controller triggers intervention;

The control module further includes a storage medium, and the storage medium includes a program capable of implementing any one of the methods described in the first aspect.

A third aspect is the application of any one of the methods described in the first aspect above in water pollution early warning.

The beneficial effects of the present invention are:

1. Compared with the existing design method, only a single state information of adjacent nodes needs to be transmitted, which reduces the communication volume in the system and reduces the communication pressure of the system. At the same time, based on the events of the system, that is, the monitoring value triggers the early warning mechanism, which reduces the energy consumption of the system;

2. When designing the convergence control protocol, that is, step 3 of this method, each node only needs to select the largest or smallest neighbor value as its own state value at the next moment, thereby obtaining a distributed event-driven consistency protocol. Compared with the traditional method, the design process of the control protocol is greatly simplified;

3. The implementation of the threshold trigger mechanism of the present invention and the early warning algorithm based on this mechanism only needs to use the information of the node and the adjacent nodes in the network that can exchange information with the node. The overall design method is based on a distributed control architecture, with It has the characteristics of low cost, strong scalability and high robustness, and is especially suitable for large-scale water source environment monitoring and early warning when a large number of distributed sensors are connected.

The present invention adopts the above-mentioned technical scheme to provide a sample document, makes up for the deficiencies of the prior art, has a reasonable design and is convenient to operate.

Description of drawings

In order to make the above and other objects, features, advantages and embodiments of the present invention more clearly understood, the accompanying drawings are described as follows:

Fig. 1 is the schematic flow chart of the present invention;

FIG. 2 is a schematic structural diagram of the wireless sensor of the present invention.

Detailed ways

Unless otherwise defined, technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The present invention employs the methods and materials described herein; however, other suitable methods and materials known in the art can also be used. The materials, methods, and examples described herein are illustrative only and not intended to be limiting. All publications, patent applications, patents, provisional applications, database entries, and other references mentioned herein are incorporated by reference in their entirety. In case of conflict, the present specification, including definitions, will control.

Example 1:

This embodiment provides a water pollution early warning system and method based on a multi-agent network convergence algorithm, including the following steps:

表示节点间的通讯边集；(i,j)∈E表示节点i可接收到来自节点j的信息；根据节点通信边关系，将整个网络拓扑构建成为一个连通图，即网络中任意一个节点出发，通过相邻节点有向边可以达到网络中任何其他一个节点。Step (1): Consider a multi-agent network composed of n agent nodes (monitoring points), denoted as G=(V,E), where V={1,2,...,n} represents the node set,

Represents the communication edge set between nodes; (i,j)∈E indicates that node i can receive information from node j; according to the node communication edge relationship, the entire network topology is constructed as a connected graph, that is, any node in the network starts from , any other node in the network can be reached through directed edges of adjacent nodes.

Step (2): According to the "Surface Water Environmental Quality Standard" (GB 3838-2002), select z water environment evaluation standards, and select the corresponding category standards according to the water area function category, and configure one or more sensors with corresponding monitoring indicators for the nodes ; Let X=[x _i,k ,k,Lng _i ,Lat _i ],k∈Z,Z={1,2,...,z}, where x _i,k represents the i-th node k-th monitoring indicators (that is, status values), Lng _i represents the longitude of node i, and Lat _i represents the latitude of node i;

The method for selecting the corresponding category standard according to the water area function category described in step (2) includes the following steps:

Step (2.1): Determine the water function category c∈C to which the water body belongs, C={c1,c2,c3,c4,c5}={Class I water, Class II water, Class III water, Class IV water, Class V water };

其中

x _c,k分别表示c类水第k个监测指标的标准上限和标准下限，c∈C；Step (2.2): Select the corresponding standard of the corresponding functional category, and set up the monitoring standard value range

in

x _{c, k} represent the upper and lower standard limits of the k-th monitoring indicator of class c water, respectively, c∈C;

x _c,k＝x _1,1＝6。In this example, the pH index of Class I water is selected as one of the monitoring and early warning states, and the specified pH value is the first index, that is, k=1; Standard basic item standard limit "pH value is in the normal threshold range of 6 to 9, then

x _c,k = x _1,1 =6.

采用分布式控制协议U＝{u₁,u₂,...,u_n}，控制协议的设计按照如下规则：节点i对t时刻收集的邻居节点j，j∈N_i发送的信息进行整理，令

则当节点监测的水域状态值超出标准区间时，预警系统启动，即当

时，或x_k,m(t)<x _k时，与之相应的控制器触发介入，具体控制器协议为

或

控制器介入后，系统中各节点的更新方程如下：Step (3): Suppose N _i (t) is the set of neighbor nodes of node i at time t,

Using the distributed control protocol U={u ₁ , u ₂ ,..., u _n }, the design of the control protocol is based on the following rules: node i organizes the information sent by neighbor nodes j, j∈N _i collected at time t ,make

Then, when the water state value monitored by the node exceeds the standard range, the early warning system is activated, that is, when

, or when x _k,m (t)< x _k , the corresponding controller triggers intervention, and the specific controller protocol is

or

After the controller intervenes, the update equation of each node in the system is as follows:

It is worth noting that the present invention only involves the information of at most one neighbor node exceeding the threshold at the same time in the system, that is, only one of the situations in which the state value of a node is higher than the upper limit of the threshold or lower than the lower limit of the threshold occurs in the system. Consider the simultaneous occurrence of both cases.

则节点不采取任何措施。而当满足

条件时，该节点判定水环境受到污染，从而触发预警，该时刻起，此后节点i将其状态值维持x_i,k(t)固定不变，并将此固定值传递给其相邻节点；网络中其余节点则按照公式(1)更新自身下一时刻的状态值，并同步节点信息X_i,k(t+1)＝[x_i,k(t+1),k，经度坐标i，纬度坐标i]，并在全网络系统传递污染坐标地址和污染的状态值，直到完成同步全部网络节点；Step (4): Node i collects and obtains the monitoring state value _xi,k (t) of its own area at time t; when the state value

Then the node does not take any action. and when satisfied

When the conditions are met, the node determines that the water environment is polluted, thereby triggering an early warning. From this moment on, node i keeps its state value x _i,k (t) fixed, and transmits this fixed value to its adjacent nodes; The remaining nodes in the network update their state values at the next moment according to formula (1), and synchronize node information Xi _,k (t+1)=[ _xi,k (t+1),k, longitude coordinate i, Latitude coordinate i], and transmit the pollution coordinate address and pollution status value in the whole network system, until the synchronization of all network nodes is completed;

The method for synchronous transmission of multi-agent network information described in step (4) includes the following steps:

Step (4.1): The node collects the data set x _i,k (t) at time t through the corresponding sensor, and obtains the latitude and longitude information through its own GPS positioning (or manually given according to the location of the node in advance);

Step (4.2): When the number monitored by node i is within the normal range, that is, it does not exceed the set threshold range, node i does not take any measures; if the state value measured by node i exceeds the threshold range, node i records An outlier at this moment, and this number is fixed for all subsequent times;

Step (4.3): After node i measures the abnormal value, it triggers the early warning mechanism, and starts to use the wireless communication protocol to send the abnormal state value to its adjacent nodes;

Step (4.4): The adjacent node j updates its own state value based on the maximum/minimum convergence algorithm, and at the same time transmits the updated state value to its adjacent nodes;

Step (4.5): Through the local information interaction, the outliers are continuously transmitted to the diffusion, until the outliers are synchronized to all nodes in the whole network to be consistent.

The conventional technology in the above-mentioned embodiment is the prior art known to those skilled in the art, so it will not be described in detail here.

While the foregoing detailed description has shown, described and indicated novel features applicable to the various embodiments, it should be understood that changes may be made in the form and detail of the illustrated apparatus or method without departing from the spirit of the present disclosure. Various omissions, substitutions and changes. In addition, the various features and methods described above may be used independently of each other or in various combinations. All possible combinations and subcombinations are intended to fall within the scope of this disclosure. Many of the above-described embodiments include similar components, and thus, these similar components are interchangeable in different embodiments. While the invention has been disclosed in the context of certain embodiments and examples, it will be understood by those skilled in the art that the invention extends beyond the specifically disclosed embodiments to other alternative embodiments and/or applications as well as their obvious Modifications and Equivalents. Therefore, it is not intended that the present invention be limited by the specific disclosure of the preferred embodiments herein.

Claims (10)

1. A water pollution early warning method based on a multi-agent network convergence algorithm is characterized by comprising the following steps:

step (1): a multi-agent network composed of n agent nodes (monitoring points), wherein the whole network topology is constructed into a connected graph according to the node communication edge relation, namely, any node in the network starts, and any other node in the network can be reached through the adjacent node directed edges;

step (2): according to the 'surface water environment quality standard' (GB3838-2002), selecting z water environment evaluation standards, and selecting corresponding classification standards according to water area function classes; configuring one or more sensors of corresponding monitoring indexes for the nodes;

and (3): suppose N_i(t) is a neighbor node set of the node i at the time t, a distributed control protocol is adopted, and the design of the control protocol is according to the following rules: the node i sorts the information sent by the neighbor node j collected at the time t, and when the water area state value monitored by the node exceeds a standard interval, the node judges that the water environment is polluted, so that early warning is triggered, and a controller corresponding to the early warning triggers intervention.

2. The method of claim 1, wherein: in step (1), the multi-agent network composed of n agent nodes (monitoring points) may be denoted as G ═ V, E, where V ═ 1,2, …, n represents a node set,

representing a communication edge set between nodes; (i, j) ∈ E indicating that node i can receive information from node j.

3. The method of claim 2, wherein: in step (2), let X ═ X_i,k，k，Lng_i，Lat_i]K ∈ Z, Z ═ 1,2_i,kIndicating the kth monitoring index (i.e., state value), Lng, of the ith node_iIndicating the longitude, Lat, of node i_iRepresenting the latitude of node i.

4. The method of claim 3, wherein: the method for selecting the corresponding category standard according to the water area function category in the step (2) comprises the following steps:

step (2.1): determining the water area function category C belonging to the water body, wherein the category C belongs to C, C is { C1, C2, C3, C4, C5}, { I type water, II type water, III type water, IV type water and V type water };

step (2.2): selecting corresponding standard corresponding to function category, and establishing monitoring standard value interval

Wherein

x_c,kRespectively representing the standard upper limit and the standard lower limit of the kth monitoring index of the C-type water, and C belongs to C.

5. The method of claim 4, wherein: in the step (3), the step (c),

neighbor node set

Distributed control protocol U ═ U₁,u₂,...,u_n}；

The control protocol is designed according to the following rules: node i collects neighbor nodes j at time t, j belongs to N_iThe transmitted information is arranged and ordered

When the water area state value monitored by the node exceeds the standard interval, the early warning system is started, namely when the water area state value monitored by the node exceeds the standard interval

When, or x_k,m(t)<x _kWhen the corresponding controller triggers the intervention, the specific controller protocol is

Or

After the controller is intervened, the state value of each node in the system is updated.

6. The method of claim 5, wherein: in the step (3), after the controller intervenes, the update equation of the state value of each node in the system is as follows:

7. the method of claim 6, wherein: further comprising the step (4): the node i acquires and acquires a monitoring state value x of a self region at the moment t_i,k(t); when state value

The node takes no action; when it is satisfied

When the node is in a condition, the node judges that the water environment is polluted, so that early warning is triggered, and from the moment, the node i maintains the state value of the node i to be x_i,k(t) is fixed and passes this fixed value to its neighboring nodes; other nodes in the network update the state value of the next moment per se according to the formula (1) and synchronize the node information X_i,k(t+1)＝[x_i,k(t +1), k, longitude coordinate i, latitude coordinate i]And transmitting the polluted coordinate address and the polluted state value in the whole network system until all network nodes are synchronized.

8. The method of claim 7, wherein: the method for synchronously transmitting the multi-agent network information in the step (4) comprises the following steps:

substeps (4.1): the node acquires a data set x at the time t through a corresponding sensor_i,k(t), obtaining longitude and latitude by self GPS positioning (or manually giving according to the position of the area where the node is positioned in advance)Information;

substeps (4.2): when the number monitored by the node i is in a normal interval, namely the number does not exceed a set threshold interval, the node i does not take any measures; if the state value measured by the node exceeds the threshold interval, the node i records the abnormal value at the moment, and then all the moments are the number fixed;

substeps (4.3): after the node i detects the abnormal value, triggering an early warning mechanism, and starting to send the abnormal state value to the adjacent node by using a wireless communication protocol;

substeps (4.4): the adjacent node j updates the state value of the adjacent node j through a maximum/minimum similarity algorithm, and transmits the updated state value to the adjacent node;

substeps (4.5): and through local information interaction, the abnormal value is continuously transmitted to the diffusion until the abnormal value is synchronized to be consistent with all nodes of the whole network.

9. A water pollution early warning system based on multi-agent network convergence algorithm is characterized by comprising:

-a monitoring module: the system comprises a multi-agent network consisting of n agent nodes (monitoring points), wherein the whole network topology is constructed into a connected graph;

-a sensor module: the system is configured at a node and can detect corresponding indexes of the corresponding node, z water environment evaluation standards are selected according to the quality standard of surface water environment (GB3838-2002), and corresponding classification standards are selected according to the function classification of a water area;

-a control module: the node comprises a distributed control protocol, wherein the protocol is used for judging that the water environment is polluted when the water area state value monitored by the node exceeds a standard interval, so that early warning is triggered, and a controller corresponding to the early warning is triggered to intervene;

the control module also comprises a storage medium, and the storage medium contains a program capable of realizing the method of any one of claims 1 to 8.

10. Use of the method of any one of claims 1 to 8 in water pollution warning.

Images