Add examples/guide without macros

[deanm0000]

The macros are great for small problems but anything a bit bigger or dynamic needs the non-macro approach. Here's a snippet of something I just did, not sure if there's a better way.

    let mut vars = ProblemVariables::new();
    let to_grid = vars.add_vector(variable().min(0.0).max(grid_out), rows);
    let to_grid_ismax = vars.add_vector(variable().binary(), rows);
    let clip = vars.add_vector(variable().min(0.0).max(max_clip), rows);

    //soc balance rule
    let lhs = charge[0] * rte[0] - discharge[0];
    constraints.push(lhs.eq(soc[0]).set_name("soc_balance_rule_0".to_string()));
    (1..rows).into_iter().for_each(|t| {
        let lhs = soc[t - 1] + charge[t] * rte[t] - discharge[t];
        constraints.push(lhs.eq(soc[t]).set_name(format!("soc_balance_rule_{}", t)));
    });
    let objective = (0..rows).into_iter().fold(Expression::from(0), |acc, t| {
        acc + to_grid[t] * prices[t] - genr[t] * ppa_price + clip[t] * ppa_price
    });
    let mut solver = vars.maximise(objective).using(default_solver);

    for c in constraints {
        solver.add_constraint(c);
    }

    let solution = solver.solve()?;

[lovasoa]

Yes, more examples are always welcome ! Feel free to contribute to the readme and to the crate doc comments

[lovasoa]

Out of curiosity, what are you working on ? This seems close to the kind of problems we work on at inthy

[deanm0000]

Given a production forecast and price forecast, trying to forecast how a battery/solar hybrid PPA holder will charge/discharge the battery in the face of the terms of their PPA. In particular they can't clip the solar except if the solar is greater than the grid connection. Additionally, they can only charge from the solar. My example above doesn't have all my constraints and variables though.

My shift to rust from pyomo is primarily motivated so I can make a polars plugin to do more Q&D battery arbitrage forecasts.

As to making a contribution, do you have any feedback on my example?

[lovasoa]

I think you could make your initial example a little bit more readable by using the constraint! macro and .sum().

• https://docs.rs/good_lp/1.10.0/good_lp/struct.Expression.html#impl-Sum%3CE%3E-for-Expression
• https://docs.rs/good_lp/1.10.0/good_lp/macro.constraint.html

[dlaehnemann]

Sorry for the random chiming in here with a question: can either of you recommend good resources or projects for battery/solar/heat pump energy managers? Sounds like you are both working in the field, and I'd be interested in open source and community-based solutions to use at home.

[lovasoa]

@dlaehnemann : I'm sorry, but I have really no idea what to recommend. My work isn't really close that; I work on proprietary software managing large industrial infrastructure

[dlaehnemann]

Cool stuff, anyways, and thanks for chiming in. Maybe the project @deanm0000 is on is more in the direction of the question.

[deanm0000]

Sorry, I work in the wholesale side of the industry too so I'm not aware of home energy managers.