Problems with modifying a driving strategy

Question

Hey there! I have a big program that needs to calculate a certain power at a waypoint. It's really really big (and of that really big part I am sure it works) but there is just this little bit that needs to be added.
To give you a bit of context:

this is part of a driving strategy for a car. It calculates the amount of work a car has to deliver when driving around the track and then tries to minimise that. That is however not the right way: it is a car that runs on hydrogen and the amount of hydrogen spent needs to be minimised. And because the efficiency of the fuel cell is not constant, the driving strategy needs a bit tweaking.

The effiency of a fuel cell can be found using a PQ diagram. It shows how much hydrogen (Q) is used when the fuel cell needs to deliver a certain power P. 
The formula for converting power P into hydrogen usage Q is this:

  Q = 79.987*P+42.942; 

That needs to be put in this part:

  function [v Wtot] = startcalculation2()
  %This function is executed when "calculate" button is pressed
  
      %   v       a nx1 matrix with all the optimal velocities [m/s]
      %   Wtot    the work at these velocities [J]
  
  %% Takes all input data from the root sea
  trackData   = getappdata(0,'trackSea');
  airData     = getappdata(0,'airSea');
  settingData = getappdata(0,'settingSea');
  calcData    = getappdata(0,'calcSea');
  
  %% Set data as input for the calculation:  
  s           =   trackData.vectors;      %[-]    ix2
  t           =   trackData.time_limit;   %[s]    1x1
  
  %% Precalculations:
  
  %Air friction precalculations:
      airData.windvec=airData.wind_dir/norm(airData.wind_dir)*airData.wind_velocity; %[m/s]
      setappdata(0,'airSea', airData);
     
      
  %% Create the figure for the converge plot:
  
      calcData.converge = 0;
      setappdata(0,'calcSea',calcData);
      %set converge count to 0:
      if settingData.converge_plot==1; 
          figure(101);
          clf;
          set(101,'Name','Converge plot work optimisation'); 
          plotconverge = axes('Parent', 101 );
          xlabel('{\itIteration count}')
          ylabel('{\itWork} (Nm)')
      else
      end
      
  
  %% OPTIMISATION:
  
  %special for lsqnonlin, lower bound:
   %lb = ones(7,1);
   
   %for j=1:trackData.number_points-2
   %    lb(j)=norm(trackData.vectors(j,:))/trackData.time_limitl
   %end
   
  %Starting point for the algorithm:
      v0=ones([trackData.number_points-2],1)*trackData.average_speed;
      %v0=ones([trackData.number_points-2],1)*1;
      
  %options for the algorithm:
      options = optimset('algorithm','trust-region-reflective','LargeScale','off');
      options.TolFun  = 1e0;
      options.TolX    = 1e0;
      options.MaxFunEvals = 1e6;
      options.MaxIter     = 1e6;
  
  % Simple optimisation:
      [x fval] = fminsearch(@calcfunction2,v0,options);
      v=x;
      
  % Simple optimisation for squares:
     %[x fval] = lsqnonlin(@xcalcfunctionINC,v0,[],[],options)
      
  % The old constraint optimisation:
     %[x, fval] = fmincon(@xcalcfunctionINC,v0,[],[],[],[],[],[],...
     %                    @xnonlinconINC,...
     %                     options);
     %v=x;
          
     
  %Calculate the last velocity variable and display result:
      n = size(s,1);
  
      trest=0;
      for j=1:(n-1)
          trest=trest + norm(s(j,:))./v(j,:);
      end
  
      v(n,:)= ( norm(s(n,:)) )/( t - trest );
      
  %total work:
      Wtot = fval;
      
  %check if time constraint is made:
      ii=trackData.number_points-1;
      
      s=zeros(ii,1);
      for j=1:ii
          s(j)=norm(trackData.vectors(j,:));
      end
      
      t=sum(s./v)
         
  
  % Plot the final velocity if its a 2 variable track:    
  if settingData.iteration_plot==1;
      fig_iter=21;
      q = figure(fig_iter);
      aa = isempty(findobj('name','2 segment track'));
      
      if aa == 1;
      else
          plot(v(1),v(2),'o','LineWidth',1,...
                      'MarkerEdgeColor','k',...
                      'MarkerFaceColor','g',...
                      'MarkerSize',10);
          
          hold off       
      end
  else
  end
  
  
  
  end

I tried adding the following:

W/t.= P
Q = 79.987*P+42.942;
Ptot=sum(P)

And then editing the minimalisation part.
That is currently this:

      

  [x fval] = fminsearch(@calcfunction2,v0,options);
        v=x;
and:

  Wtot = fval;

So in short:
The program in its current form minimises the amount of Work. That is wrong. It should minimise the amount of hydrogen spent.
To do this I need to add the following:
Work / time = power
power in the formula to get the amount of hydrogen/per time
Hydrogen per time * time = hydrogen
minimise the hydrogen.

How to add this?

Problems with modifying a driving strategy

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

回答（0 个）

另请参阅

类别

标签

Community Treasure Hunt

Problems with modifying a driving strategy

0 个评论 显示 -2更早的评论隐藏 -2更早的评论

回答（0 个）

另请参阅

类别

标签

Community Treasure Hunt

0 个评论
显示 -2更早的评论隐藏 -2更早的评论