Since the end of last century, there have been quick and important changes in the field of regional traortation and developing number of city cars has increased traffic problems in most metropolitan areas. Urban traortation planning is an unavoidable factor due to constraints on construction budgets and high costs of urban traortation projects. Thus, drawing a traffic pattern between different regions via the travel demand matrix is one of the most important requirements for traortation planning. Using traditional methods such as interviewing or distributing questionnaires to gather information about the travel demand matrix is one of the costliest steps in traortation planning. Traffic sensors are new approach in the gathering of traffic information in which attract many attentions of researchers and traortation planners. Automated Vehicle Location (AVL) is one of the advanced sensors which enables to locate vehicles equipped with Global Positioning System (GPS). Automated Vehicle Location detects the vehicle's location and determines its path when the vehicle with GPS is turned on. While this sensor can provide valuable traffic information to traortation planners, but, it has problem with determining the GPS matrix of travel due to incomplete information of equipping only a portion of the vehicles currently available from the AVL sensor. In this thesis, a proposed model is presented to estimate the travel demand matrix with the collected information of GPS-based AVL sensors and traffic volume sensors in the network.