Evaluation Index and Evaluation Method of Three-Phase Imbalance Treatment Effect Based on Commutation

With the construction and development of the distribution Internet of things in China, the three-phase imbalance treatment in the distribution area has gradually attracted attention as an optional function, and a large number of intelligent terminal manufacturers have begun to study this aspect. In this paper, research is carried out on the three-phase imbalance treatment function based on commutation control. The evaluation index system of imbalance treatment based on commutation is proposed in combination with the imbalance harm and the influence of treatment, and the three-phase imbalance treatment effect is evaluated by combining the analytic hierarchy process and improved entropy weight method. An example is given to verify the effectiveness of the evaluation method proposed in this paper.

At present, the methods to deal with three-phase imbal-29 ance mainly include load phase sequence balance, distri-30 bution network reconfiguration, and three-phase imbalance 31 compensation devices. Each treatment method has its own 32 advantages and disadvantages, and the scope of application 33 is also different [3], [4], [5], [6], [7], [8], [9]. The imbalance 34 treatment method based on commutation switch collects the 35 data of the terminal by analyzing the electricity consumption 36 information, the optimal control strategy is used to switch the 37 phases of a certain number of single-phase users, and the load 38 redistribution on the three-phase line is realized [8], [9].  Aiming at the problem that there is, at present, neither an 101 evaluation index nor an evaluation method for three-phase 102 imbalance treatment based on commutation, this paper starts 103 from the imbalance influence and the imbalance treatment 104 mechanism of load balancing, proposing an evaluation index 105 system for three-phase imbalance treatment based on com-106 mutation switch. The subjective weight of each index is given 107 by the analytic hierarchy process, and the objective weight 108 of each index is determined by the improved entropy weight 109 method while overcoming the defect of polarizing index 110 importance. The three-phase imbalance treatment based on 111 commutation is comprehensively evaluated by combining 112 subjectivity and objectivity. When a serious three-phase imbalance occurs in the outlet 119 of the distribution area, it increases the loss of the power 120 grid, endangers the electrical equipment and lines, causes the 121 malfunction of the protection device, brings economic losses 122 to the power supply enterprise, and affects the power quality 123 of the users. The root cause of its harm is mainly reflected in 124 the increase in current loss and voltage deviation caused by 125 imbalance.

126
The increase in current loss caused by imbalance includes 127 the load loss of distribution transformers in the station area, 128 the loss of low-voltage lines, and the loss of high-voltage 129 lines in the station area. The increase in loss not only directly 130 affects the economic benefits of power supply enterprises but 131 also causes the superheating of equipment and lines. It brings 132 about problems such as equipment damage and line failure, 133 which indirectly affect the safety and economy of the power 134 supply. The factor that affects the degree of loss is the current 135 size of each phase. Therefore, not only the degree of imbal-136 ance but also the size of the current, that is, the loading rate, 137 should be considered when performing imbalance treatment. 138 The three-phase load imbalance causes voltage deviation 139 in each phase, which causes a certain degree of damage to 140 the high-and low-voltage equipment. The calculation of the 141 voltage deviation is not related to the size of the current but 142 only to the imbalance degree of the current.

143
Therefore, the treatment effect index must be comprehen-144 sively considered from the two aspects of imbalance degree 145 and loading rate. The regulation benefit is defined as a first-level index of 148 three-phase imbalance treatment, reflecting the benefits that 149 can be brought after treatment. There are two second-level 150 indices, namely, the imbalance degree and the imbalance loss 151 coefficient.

152
Imbalance degree, k: The imbalance degree of the three-153 phase current, according to the State Grid Corporation's 154 enterprise standard (Q/GD W519-2010) ''Procedure for dis-155 tribution network'', can be calculated as In the formula, I max is the significant value of the maximum current in three phases, and I min is the significant value of the 159 minimum current in three phases.

160
The imbalance loss coefficient, α, the product of the load-161 ing rate, and the imbalance degree reflect the influence of the 162 current imbalance on the loss.
In the formula, β is the loading rate of the transformer.

165
The overall loss caused by imbalance is a function of the Although the ''Notice'' sets out the regulations for one over-224 limit day (continuous overlimit time for more than 1 hour in 225 one day) and the conditions that the distribution area needs 226 to be treated, nonetheless it does not specify what standards 227 must be met after treatment. Therefore, various treatment 228 strategies can easily achieve this index, such as taking action 229 every 45 minutes to keep it from exceeding the limit continu-230 ously for 1 hour, but it cannot reflect the ultimate reduction in 231 loss and the improvement in voltage quality. Longer treatment 232 intervals (one action every 45 minutes) may not be able to 233 guarantee the load balance state after treatment, while shorter 234 treatment intervals (one action every 15 minutes) will bring 235 about the frequent action of the commutation switch. There-236 fore, the three-phase imbalance treatment evaluation not only 237 considers the treatment effect under the condition of a fixed 238 load but should also consider the comprehensive treatment 239 effect over a period of time.

240
Considering the treatment effect and the influence intro-241 duced in the treatment process, this paper divides the three-242 phase imbalance treatment index system into two first-level 243 indices: Regulation benefit and regulation cost. The two first-244 level indices correspond to the two second-level indices. The 245 treatment index system is shown in Figure 2.  Table 1. Because it is a periodic eval-250 uation, the specific value of the imbalance influence index 251 after treatment is the statistical value of the time occupied by 252 each level.

253
The second-level indices of regulation cost are also divided 254 into five levels, and the threshold value of each level is deter-255 mined through specific test cases. A three-phase imbalance 256 test scheme is designed, in which six fixed loads and six 257 adjustable loads can be set, each load has an adjustment range 258 of 0-3 KW, and the adjustment step value is 0.1 KW. Given 259 the load test data for a period of time, calculate all possible 260 switch action combinations, count the action combinations 261 with an imbalance loss coefficient greater than level 5 after 262 adjustment, and then determine the threshold value of each 263 VOLUME 10, 2022     Information entropy is a measure of the uncertainty of a 319 random event. The basic idea of the entropy weight method 320 for determining the objective weight is as follows: according 321 to the test data, the entropy value e i of each index of the three-322 phase imbalance treatment is calculated and compared, and 323 then the corresponding weight of each index is determined 324 according to the entropy value. The specific steps are as 325 follows.

327
The treatment effect index set V = {v 1 , v 2 . . . , v n } is a set 328 composed of n indices to evaluate the treatment effect, and 329 the evaluation set Q = {q 1 , q 2 . . . , q m } is a set composed of 330 m evaluation results of each evaluation index. Select excel-331 lent, fine, passed, poor, or very poor, corresponding to the 332 1st to 5th levels of the evaluation results, respectively. The 333 evaluation function, f ij , from index v i to evaluation element 334 q j is established, and then the fuzzy evaluation matrix F from 335 V to Q is obtained, which is expressed as follows.

338
The entropy value of each index is calculated by the improved 339 entropy weight method. First, for the judgment matrix, the 340 entropy value, e i , of the relative importance of a certain index, 341 v i , is calculated. To prevent the variation of some index data from affecting 344 the weight of the corresponding index, this paper uses acti-345 vation function logistics to improve the traditional entropy 346 weight method, and the improved entropy value is e i .

359
The evaluation process of the three-phase imbalance treat- Perform a weighted averaging on R to obtain the final 387 evaluation result D.

390
According to the three-phase imbalance test method 391 described in Section III, C, the load test data are given, the 392 total time is 2 hours, the load changes every 15 minutes, 393 and the level limit of each index is obtained according to the 394 load data. The factor set V and the judgment set Q of the 395   According to the load data, the imbalance loss coefficient and 421 the time when the imbalance degree is at different levels after 422 the imbalance treatment of the tested intelligent terminal are 423 counted. Table 4 is a set of statistical results.

424
The probability occupied by each level is calculated 425 according to Formula (10), and together with the statistical 426 value of the switch action index, a judgment matrix F is 427 formed, as shown in Formula (14).

481
This paper focuses on the three-phase imbalance treatment 482 evaluation of intelligent terminals in the distribution area 483 based on commutation switching and performs the following 484 work:

485
The influence of the three-phase imbalance in the distribu-486 tion area and the three-phase imbalance treatment principle 487 based on the commutation switch are analyzed, the treatment 488 effect and the influence introduced by the treatment process 489 are combined, and the index system for the three-phase imbal-490 ance treatment of an intelligent terminal is proposed.

491
The weight of each evaluation index is determined by using 492 AHP, and the improved entropy weight method is used to 493 overcome the influence of some index data concentrations 494 and determine the objective weight. The subjective weight 495 and objective weight are combined to evaluate the three-phase 496 imbalance treatment effect of the intelligent terminal.

497
The effectiveness of the evaluation method is verified by a 498 set of examples. However, the algorithm still has some short-499 comings, which cannot accurately distinguish the number of 500 switch actions, and the evaluation results will be significantly 501 different when its value is near the classification threshold 502 value.