Estimation of Kinetic Parameters from Surface Plasmon Resonance Sensorgrams

This paper proposes the use of the Recursive Least Squares (RLS) method for the estimation of kinetic adsorption constants in Surface Plasmon Resonance (SPR) biosensor experiments. The RLS method was chosen for this application due to its real-time adaptive estimation capabilities, allowing efficient parameter updates as new data become available. The proposed method was applied using the DOCE SPR sensor platform to obtain consistent kinetic parameters for different molecules, including NaOH, PBS, and HCl, across various concentrations, treating each as a distinct input to derive specific kinetic constants. The results indicate that the estimated kinetic constants for different concentrations of the same substance are closely related, demonstrating the robustness of the RLS method. The relatively low least-squares error further validates the accuracy of our model fits. The efficiency of the implemented estimation method was demonstrated through performance metrics, with an average iteration time of 0.012053s per experiment for PBS. In summary, the results corroborate the effectiveness of the RLS method for real-time kinetic constant estimation and highlight the potential for future work in exploring more robust kinetic models, such as non-linear models and neural networks, to capture the logistic tendencies observed in the dataset.