Congestive Heart Failure (CHF) is a life-threatening cardiovascular condition that requires early detection and precise diagnosis to enhance patient outcomes. The aim of this study is to provide a fully automatic system to improve the process of diagnosing and classifying CHF from other Cardiac Arrhythmia (ARR) and Normal Sinus Rhythms (NSR). This study introduces a novel approach that combines machine learning techniques with the Grasshopper Optimization Algorithm (GOA) to enhance classification accuracy while reducing computational complexity for detecting and classifying ECG signals. The methodology integrates feature extraction and optimization method with GOA to enhance classification accuracy while reducing computational complexity. In order to evaluate the effectiveness and performance of the proposed method, 162 recordings of ECG signals for cases with ARR, CHF and NSR from three open access databases (MIT-BIH ARR, MIT-BIH NSR, and BIDMC for CHF data) were used. Machine learning models including support vector machines (SVM) and Random Forest (RF) methods are applied to analyze ECG signals and distinguish CHF patients from NSR and ARR. For this purpose, chaotic features like recurrence plot was extracted as a two-dimensional map from ECG signals related to Heart Rate Variability (HRV) and ECG Derived Respiration signals. HRV and ECG Derived Respiration signals are extracted from ECG signal and then by Recurrence Plot diagram, these signals are converted into two-dimensional images for using machine learning methods. After extracting high-level features from graphs related to HRV and ECG Derived Respiration in parallel, we optimized SVM parameters using GOA for increasing classification accuracy. The results show that the method proposed in this article with an overall accuracy of above 98% has performed better than other previous related researches and has several advantages. The implementation of this algorithm in order to diagnose heart disease shows the significant application of AI in medicine and early diagnosis.