ORIGINAL ARTICLE ‘‘Self-organizing maps’’ for identification of tire model longitudinal braking parameters of a vehicle on a roller brake tester and on flat ground C. Senabre • E. Velasco • S. Valero Received: 8 November 2010 / Accepted: 14 June 2011 / Published online: 8 July 2011 Ó Springer-Verlag London Limited 2011 Abstract This paper discusses how to identify tire model coefficients that are used to compare longitudinal forces when braking. Those in the automotive world have worked extre- mely hard to obtain these parameters and different methods have been used to match the values of these parameters. This paper proposes the use of self-organizing maps to tackle this problem whereby interactively searches are carried out to find the optimum tire model parameters. The objective of this research is to prove the capability of self-organizing maps (SOMs) to classify a vehicle’s braking formula on a roller brake tester from the MOT (Ministry of Transport) and on flat ground. The neural network produced a good brake-slip ratio when presented with data that are not used in network training. This means that the methodology is feasible. This tool easily obtains the brake-slip equation of each experiment and the braking on two different experimental tests will be compared. Keywords Self-organizing maps  Tire model  Parameters identification 1 Introduction When a vehicle is taken to the (Ministry of Transport) MOT testing facilities, the roadworthy inspection includes a brake test carried out on the rollers to check the brake circuit. Several questions on how efficient the MOT testing facilities are need to be answered: Does braking on a roller brake tester accurately reproduce braking on flat ground? To what extent does tire pressure affect the measurements taken on the rollers? Is this test safe enough to assess the condition of brakes? Is the brake test 100% effective? The aim of the study is to calculate a vehicle’s braking capacity by measuring slippage on a roller brake tester at MOT centres, compare these with similar measurements taken on flat ground, and use the result to assess the machine’s reliability to test brake systems. Detailed knowledge of each brake system is required to actually be able to contribute anything worthwhile to these programs which is obtained through ‘‘The Pacejka96 for- mula’’ [1] on experimental data and the characterization of these brake clusters. This high-dimensional data set cannot be modeled eas- ily, and advanced tools are needed to synthesize structures from this information, for instance, data mining research and applications [2]. This study aims to evaluate self- organizing maps (SOMs) to obtain ‘‘The Pacejka96 for- mula’’ of each brake on the roller tester. The main objective of this paper is to provide the results of training a neural network using field data to predict the brake-slip ratio. 2 Background In 1936, Alan Turing was the first person to study the brain as a way of seeing the world of computing. It was not until 1943 that W. McCulloch, a neurophysiologist, and W. Pitts, a mathematician, devised a theory about how neurons work and laid the foundations of neural computation. The C. Senabre (&)  E. Velasco  S. Valero Universidad Miguel Herna ´ndez de Elche, Dpto. de Ingenierı ´a Meca ´nica y Energı ´a, Avd. de la Universidad s/n. Edificio Quorum-V, Elche, Spain e-mail: csenabre@umh.es E. Velasco e-mail: emilio.velasco@umh.es S. Valero e-mail: svalero@umh.es 123 Neural Comput & Applic (2012) 21:1775–1782 DOI 10.1007/s00521-011-0666-7