I strongly doubt that you could get an AI to estimate a specific model for you. AI's can only address topics on which they have been trained, so even performing complex addition is often not possible, because the the AI has never seen the set of numbers in question.

You simply need to start by writing your equations and use them to define a reinforcement learning model. Train a reinforcement learning network using a portion of the data and evaluate it with another portion.

Note that the RL agent's output model and the proper definition of the reward model (to achieve the desired objective) are essential.

Start with clean data, visualize it using graphical representations, and then use Machine Learning techniques to apply piecewise functions or non-linear models for fitting the relationships effectively. That's it. Feel free to ask for the tips on https://wa.me/+923440907874.