Coding up basic power flow equation

Specific Domains Optimization (Mathematical)
,
5

Thanks for the response @odow.
Since, I’m unable to upload the csv files here, I have made the simplified data available on my github page: https://github.com/m-bah/test

A slightly more expanded version of my coding exercise;

using JuMP, Clp, DataFrames, CSV
using Base: Symbol

function df_to_dict_with_id(df::DataFrame)
  Dict(col[1] => Dict(zip(df[!, 1], col[2])) for col in pairs(eachcol(df)))
end

#Read in data

data_dir = pwd()*"/test/"
data = Dict{Symbol,Any}()
set = Dict{Symbol,Any}() 
readDataDFs = Dict()
inputDataItems = ["plants", "demand", "nodes","lines"]

for inputDataItem in inputDataItems #loop starts~ for each data file inputDataItems
  readDataDFs[Symbol(inputDataItem)] = DataFrame(CSV.File(data_dir * inputDataItem * ".csv"), copycols=true) #read in all the CSV files into the dictionary readDataDFs
  data[Symbol(inputDataItem)] = df_to_dict_with_id(readDataDFs[Symbol(inputDataItem)]) 
  set[Symbol(inputDataItem)] = readDataDFs[Symbol(inputDataItem)].index 
end

P = set[:plants] #index of all plants
C = filter(r -> any(occursin.(["conventional"], r.plant_type)), readDataDFs[:plants]).index #index of conventional plants
N = set[:nodes] #Index of nodes  
T = set[:demand] #Time stamp
L = set[:lines] #Index of power lines

basic_model = Model(Clp.Optimizer) 

#Define variables
@variable(basic_model, GEN[T, C] >=0) #Conventional plant generation
@variable(basic_model, FLOW[T, L] ) #power flow

@objective(basic_model, 
  Min, 
  sum(sum(GEN[t, p]*data[:plants][:cost][p] for p in C) for t in T)) 

@constraint(basic_model, 
  CapConstraint[t in T, c in C],
  GEN[t,c] <= data[:plants][:g_max][c] ) # Capacity constraint

.... #steps skipped

## issue starts here

flow1 = sum(1/3*GEN[1:3,"basel"] - 1/3*GEN[1:3,"bern"])
flow2 = sum(2/3*GEN[1:3,"basel"] + 1/3*GEN[1:3,"bern"])
flow3 = sum(1/3*GEN[1:3,"basel"] + 2/3*GEN[1:3,"bern"])

@constraint(basic_model, 
    LineFlowA[t in T, l in L],
    FLOW[t,l] == flow1 )

@constraint(basic_model, 
    LineFlowB[t in T, l in L],
    FLOW[t,l] == flow2 )

@constraint(basic_model, 
    LineFlowC[t in T, l in L],
    FLOW[t,l] == flow3 )

@constraint(basic_model, 
    FlowMax_Pos[t in T, l in L],
    FLOW[t,l] <= data[:lines][:f_max][l]
    ) #positve line flow  constraint

@constraint(basic_model, 
    FlowMax_Neg[t in T, l in L],
    FLOW[t,l] >= -data[:lines][:f_max][l]
    ) #negative line flow constraint

When I try to compute flow1 for instance, i get the following error message.

julia> flow1 = sum(1/3*GEN[1:3,"basel"] - 1/3*GEN[1:3,"bern"])
ERROR: MethodError: no method matching promote_shape(::Tuple{Vector{Int64}}, ::Tuple{Vector{Int64}})
Stacktrace:
 [1] promote_shape(a::JuMP.Containers.DenseAxisArray{AffExpr, 1, Tuple{Vector{Int64}}, Tuple{JuMP.Containers._AxisLookup{Dict{Int64, Int64}}}}, b::JuMP.Containers.DenseAxisArray{AffExpr, 1, Tuple{Vector{Int64}}, Tuple{JuMP.Containers._AxisLookup{Dict{Int64, Int64}}}})
   @ Base .\indices.jl:169
 [2] -(A::JuMP.Containers.DenseAxisArray{AffExpr, 1, Tuple{Vector{Int64}}, Tuple{JuMP.Containers._AxisLookup{Dict{Int64, Int64}}}}, B::JuMP.Containers.DenseAxisArray{AffExpr, 1, Tuple{Vector{Int64}}, Tuple{JuMP.Containers._AxisLookup{Dict{Int64, Int64}}}})
   @ Base .\arraymath.jl:38
 [3] top-level scope
   @ REPL[26]:1

In short, my main aim is to manually code up equality constraints for individual line flows (i.e. LineFlowA - LineFlowC).

Apologies for the long and messy response

Thanks