This paper presents generalised particle approach (GPA) to optimise bandwidth allocation and QoS in ATM networks. The generalised particle approach transforms the network bandwidth allocation problem into kinematics and dynamics of numerous particles in a force field. ATM traffic is bursty in nature; the dynamic bandwidth allocation is an important factor for network performance. The network utilisation is reasonably attained by dynamically allocating the bandwidth. In this approach of ATM network optimisation, GPA deals with variety of phenomena like congestion, failure and interaction. The payment policy of ATM network path determines the bandwidth allocation of network link. It changes dynamically to the situation of demands and supplies of bandwidth. The path particle and link particle move in their respective force fields and the bandwidth allocation is done through an iterative procedure. The algorithm is simulated using the MATLAB. The performance of network based on GPA is compared with the max-min fairness (MMF) algorithm. The GPA algorithm based control of network has better network bandwidth utilisation, success rate and price allocation.