Cell Balancing (CB) is a crucial aspect of the Battery Management System (BMS), which is used to increase the batteries operate time as well as operational life. The most widely used method is Passive Cell Balancing (PCB) since it is inexpensive and simple to use. This work proposes an algorithm using Deep Learning (DL) that selects the balancing resistor effectively with respect to increasing temperature, C-rate, level of cell imbalance, and balancing duration. Convolution Neural Network (CNN), hybrid method of Convolution Neural Network (CNN)-Long short-term memory (LSTM) (CNN-LSTM) and Convolution Bidirectional LSTM (CNN-BiLSTM) are used to assess the performance of the suggested system. In order to optimize the balancing parameters, the balancing system's error analysis is carried out, and performance indices like Mean Square Error (MSE), Root Mean Square Error (RMSE), and Mean Absolute Error (MAE) are used to compare the suggested algorithms. The hybrid CNN-BiLSTM obtained MAE of 0.0453, MSE of 0.0062 and RMSE value of 0.0671 which is better than other models.