Internet of Things Detection and Location Based on Remote Communication Fault of Power Consumption Acquisition Module

The access of distributed generation makes the single-source radial distribution network become a complex multi-source network. The power flow, network loss, voltage, and so on are also changing. At the same time, it aggravates the complexity of remote communication fault location of remote communication fault power acquisition module based on the Internet of Things. In this paper, a new switching function and evaluation function are used to adapt to the situation of distributed generation access. Considering that the fault section location is a discontinuous problem, the binary particle swarm algorithm is used to make full use of its fast convergence and combine with the operation of diversity preservation to realize the fault section location. The experimental results show that the hybrid algorithm in this paper has a good effect on fault recovery, and has certain advantages and theoretical research value in optimization ability.


Introduction
Under the background of energy-saving and environmental protection, distributed generation technology with less pollution and high flexibility has been developed rapidly. At the same time, it also has an inevitable impact on the operation mode of the traditional distribution networks. In the distribution network with distributed generation, how to quickly and accurately locate the fault section of the distribution line and restore the fault has become a key link in the safe and reliable operation of electric power.
In reference [2], the incomplete or distorted fault information is taken into consideration in the fault section location. The conformity of each section to the fault judgment basis is checked one by one for single fault and multiple fault scenarios under multi-source scenarios. Reference [3] omits the reverse correlation nodes when constructing the network description matrix, which enhances the sparsity of the matrix. The lack of a simple judgment basis is the common deficiency of literature [4] and literature [5], which aggravates the complexity of the judgment process. In reference [6], the author uses the short-circuit current amplitude at the switch to form a fault information matrix, which is combined with Kirchhoff's current law to judge, and gives the corresponding strategy when the fault information is not fully obtained.
The integration of distributed generation (DG) makes the voltage and fault current flow direction of the distribution network change. Thus, it is necessary to fully consider the influence of DG in the construction of switching function and the selection of fault section location method. The branch 2 structure of the distribution network is relatively complex. When the direct algorithm is used to locate the section, the amount of data information is large and the fault tolerance is deficient. The indirect algorithm, namely the artificial intelligence algorithm, has become the main method of fault section location research, which has a good fit with the operation of the computer.
The innovation of this paper is to make full use of the fast convergence of particle swarm algorithm and the crossover, mutation, and other operations of genetic algorithm, which are conducive to maintaining diversity, complementing each other, and achieving the location of fault sections together. Finally, the fault section location test is carried out in the distribution network with distributed generation, and the results show that the proposed algorithm has better performance.

Remote communication fault location model of power consumption acquisition module
After the distributed generation is connected, the power flow in the distribution network changes and the coding mode, switching function, and fitness function need to be improved to adapt to the changes brought by the distributed generation [7]. The remote communication fault location model of the power consumption acquisition module is shown in Figure 1. (1) Coding rules The current direction in the traditional distribution network is in the positive direction from the power supply to the user by default. However, after the distributed generation is connected, the power flow direction of the distribution network changes and the traditional 0-1 coding can no longer solve the actual problem. At this time, a new coding method needs to be developed to specify that the current flow direction in the system is in the positive direction under normal conditions, as shown in Formula 1: Where: I j is the state of the node, and the value is ( ) The status code value of each switch node in case of remote communication failure of power acquisition module containing DG is shown in Table 1.  Table 1 Status code values for each switch node (2) Switch function modification After the distributed power supply is connected, the previous switching function is used. Each power supply needs to be analyzed as the main power supply, and the efficiency is reduced by multiple assumptions and multiple positioning in the positive direction. Therefore, in this paper, the switch node functions transformed by the following formulas (2) to (4) are used to locate the remote communication fault section of the power acquisition module connected to the distributed generation [8]. It is specified that the section from the switch node to the main power supply is the upstream section. The section from the switch node to the user (or DG) is the downstream section.
represents an upstream switch node function. represents a downstream switch node function. is a final switch node function.  Table 2.  When locating the fault section, the fitness function used is often misjudged and can not achieve the desired effect, so the modified switching function module [10] is used.
λ is the weight coefficient in the "minimum set" of the diagnosis theory. Its value range belongs to [0, 1], and its value is usually 0.5. It can be seen from the formula that the smaller the fitness value is, the better it is.

Experimental environment
Faults can be divided into different types according to different criteria. To verify the feasibility of the hybrid algorithm in fault section location, three aspects are considered, including no DG access, single DG access and multiple DG access. Simulations of single fault, multiple faults, with information distortion and without information distortion are set up in different branches. It is a more direct and effective proof of the effectiveness of the hybrid algorithm in this paper to set a comprehensive fault in a variety of situations. Take the topological structure system shown in Figure 2 as an example to set and verify the faults in various situations.

Parameter setting
Write the code of the hybrid algorithm in MATLAB and set the relevant parameters of the algorithm, as shown in Table 3 and Table 4 below.

Analysis of experimental results
When the distributed power source DG1 is connected to the distribution network, a three-phase short circuit fault is set in the branch section 28 containing the distributed power source. The optimal solution of the fitness function can be obtained by inputting the switch state information:  Table 5. The results prove that the hybrid algorithm can accurately locate the fault section under single fault.

Conclusion
In this paper, the improved particle swarm optimization algorithm is combined with the genetic algorithm. To solve the discrete problem of locating the remote communication fault section of the power consumption acquisition module, the improved algorithm is used to improve the individual diversity by making full use of the convergence of the particle swarm and the crossover, mutation, and other operational factors of the genetic algorithm. Finally, through the simulation of different distributed generation access and different faults, it is verified that the two algorithms complement each other and have good fault tolerance, so they can quickly and stably achieve the experimental results of local optimization and global optimization.