This article presents a method for approximating helicopter turboshaft engine parameters using a neural network to improve efficiency. The approach involves a three-layer direct propagation neural network with linear neurons in the output layer, trained using a modified scaled conjugate gradient algorithm. This modification introduces a moment coefficient to avoid local minima during training. The method enables the determination of the optimal gas-generator rotor r.p.m. region for a specific type of helicopter turboshaft engine, thereby optimizing energy consumption and compressor efficiency. Experimental data show that the proposed method achieves an efficiency coefficient of 0.994, which is significantly higher than that of traditional methods. The neural network is trained using data from helicopter flights, and its performance is evaluated based on root mean square error, mean absolute error, and mean absolute percentage error. The results indicate high accuracy in approximating helicopter turboshaft engine parameters, with maximum errors not exceeding 1.8%. The study also demonstrates the effectiveness of the modified training algorithm in avoiding local minima and improving the neural network's ability to generalize and approximate parameters. The results show that the neural network can effectively model the relationships between engine parameters and gas-generator rotor r.p.m., enabling more efficient and reliable helicopter turboshaft engine operation. The study contributes to the development of control technologies and end-use energy in helicopter turboshaft engine operations, helping to optimize engine performance and reduce fuel consumption.This article presents a method for approximating helicopter turboshaft engine parameters using a neural network to improve efficiency. The approach involves a three-layer direct propagation neural network with linear neurons in the output layer, trained using a modified scaled conjugate gradient algorithm. This modification introduces a moment coefficient to avoid local minima during training. The method enables the determination of the optimal gas-generator rotor r.p.m. region for a specific type of helicopter turboshaft engine, thereby optimizing energy consumption and compressor efficiency. Experimental data show that the proposed method achieves an efficiency coefficient of 0.994, which is significantly higher than that of traditional methods. The neural network is trained using data from helicopter flights, and its performance is evaluated based on root mean square error, mean absolute error, and mean absolute percentage error. The results indicate high accuracy in approximating helicopter turboshaft engine parameters, with maximum errors not exceeding 1.8%. The study also demonstrates the effectiveness of the modified training algorithm in avoiding local minima and improving the neural network's ability to generalize and approximate parameters. The results show that the neural network can effectively model the relationships between engine parameters and gas-generator rotor r.p.m., enabling more efficient and reliable helicopter turboshaft engine operation. The study contributes to the development of control technologies and end-use energy in helicopter turboshaft engine operations, helping to optimize engine performance and reduce fuel consumption.