Hybrid Lithium-Ion SOC Model MATLAB
Budget: ₹12,500 – ₹37,500 INR
I have a set of logged drive-cycle data from a lithium-ion pack and need a robust hybrid State-of-Charge estimation model built in MATLAB. My goal is to blend state-of-the-art algorithms—think extended/unscented Kalman, adaptive observers or similar—with a physics-based cell model (probably equivalent circuit model)so the estimator performs reliably across varying temperatures, C-rates and ageing conditions.
What I already have
• Raw current, voltage and temperature traces for different operating scenarios
• Basic cell specs and key parameters (capacity, R0/R1, OCV table)
What I still need
• A clean, well-structured MATLAB codebase (functions or Simulink blocks) that:
– Implements the chosen hybrid algorithm
– Encapsulates an accurate lithium-ion equivalent-circuit or electrochemical model
– Runs batch simulations against my data, producing SOC, error metrics and plots
Acceptance criteria
1. SOC root-mean-square error ≤ 2 % on the provided data sets.
2. Modular code with clear comments so I can update parameters or switch algorithms later.
3. Short write-up (PDF or Markdown) detailing the model assumptions, tuning steps and how to extend it to new cells.
MATLAB/Simulink, Optimization Toolbox or Battery Toolbox can be used—whichever best fits your approach. Once delivered, I’ll run the estimator on an unseen data set and, if the accuracy holds, we’re done.
What I already have
• Raw current, voltage and temperature traces for different operating scenarios
• Basic cell specs and key parameters (capacity, R0/R1, OCV table)
What I still need
• A clean, well-structured MATLAB codebase (functions or Simulink blocks) that:
– Implements the chosen hybrid algorithm
– Encapsulates an accurate lithium-ion equivalent-circuit or electrochemical model
– Runs batch simulations against my data, producing SOC, error metrics and plots
Acceptance criteria
1. SOC root-mean-square error ≤ 2 % on the provided data sets.
2. Modular code with clear comments so I can update parameters or switch algorithms later.
3. Short write-up (PDF or Markdown) detailing the model assumptions, tuning steps and how to extend it to new cells.
MATLAB/Simulink, Optimization Toolbox or Battery Toolbox can be used—whichever best fits your approach. Once delivered, I’ll run the estimator on an unseen data set and, if the accuracy holds, we’re done.