Position estimation techniques for solenoid actuators are successfully used in a wide field of applications requiring monitoring functionality without the need for additional sensors. Most techniques, which also include standstill condition, are based on the identification of the differential inductance, a parameter that exhibits high sensitivity towards position variations. The differential inductance of some actuators shows a non-monotonic dependency over the position. This leads to ambiguities in position estimation. Nevertheless, a unique position estimation in standstill condition without prior knowledge of the actuator state is highly desired. In this work, the eddy current losses inside the actuator are identified in terms of a parallel resistor and are exploited in order to solve the ambiguities in position estimation. Compared to other state-of-the-art techniques, the differential inductance and the parallel resistance are estimated online by approaches requiring low implementation and computation effort. Furthermore, a data fusion algorithm for position estimation based on a neural network is proposed. Experimental results involving a use case scenario of an end-position detection for a switching solenoid actuator prove the uniqueness, the precision and the high signal-to-noise ratio of the obtained position estimate. The proposed approach therefore allows the unique estimation of the actuator position including standstill condition suitable for low-cost applications demanding low implementation effort.