The development of highly advanced technology allows for the transmission of data without media or what is commonly called wireless (wireless) including sending power commonly called Wireless Power Transfer (WPT). Many factors are known in power delivery at WPT, one of which is the coupling coefficient (k) value. So far, the k-value approach has been carried out by entering values between 0 and 1, making this WPT research inefficient. Therefore, the solution that researchers can provide is to design a coupling coefficient detection algorithm based on frequency response with the Manhattan distance metric. This algorithm will detect the coupling coefficient value based on the frequency and power values at the input. The output result of this algorithm is the corresponding coupling coefficient value. The dataset used in this study consisted of 19 classes with each class totaling 1000 data so that the total dataset was 19000 data. The test results of this study are based on test data, namely getting an accuracy of 95% for coupling coefficient detection, the average computational time required for testing is 0.0274 seconds, and the percentage of error rate from the comparison of the algorithm with the simulation results is 0%..
Copyrights © 2022