Converged to an infeasible point.

Hello,

can someone help me with this problem?

Thats my optimization problem

if f_lag > 0                % decision braking or driving mode
       fun =@(x) (A/SOC_1)*abs(x(1))+(A/SOC_2)*abs(x(2))+(A/SOC_3)*abs(x(3))+(A/SOC_4)*abs(x(4))+d_delay_1-(SW_1+x(5))+d_delay_2-(SW_2+x(6))+d_delay_3-(SW_3+x(7))+d_delay_4-(SW_4+x(8));
 else
      fun =@(x) ((B*M_reg_1)/SOC_1)*(1/x(1))+((B*M_reg_2)/SOC_2)*(1/x(2))+((B*M_reg_3)/SOC_3)*(1/x(3))+((B*M_reg_4)/SOC_4)*(1/x(4))+d_delay_1-(SW_1+x(5))+d_delay_2-(SW_2+x(6))+d_delay_3-(SW_3+x(7))+d_delay_4-(SW_4+x(8));
end

x0 = [1,1,1,1,1,1,1,1];
A = [];
b = [];
Aeq = [0 0 0 0 1 0 0 0; 0 0 0 0 0 1 0 0; 0 0 0 0 0 0 1 0; 0 0 0 0 0 0 0 1; 0 0 0 0 0 0 0 0; 0 0 0 0 0 0 0 0];

SOC_dif1=SOC_1-(SOC_1+SOC_2+SOC_3+SOC_4);
SOC_dif2=SOC_2-(SOC_1+SOC_2+SOC_3+SOC_4);
SOC_dif3=SOC_3-(SOC_1+SOC_2+SOC_3+SOC_4);
SOC_dif4=SOC_4-(SOC_1+SOC_2+SOC_3+SOC_4); 
if f_lag==1
     
    if (SOC_dif1 < 0.15)          % state of charge of one actor is really low, actor is only used for steering %during driving mode
                                                        %SOC_krit = 0.15
        Aeq(6,1) = 1;
    end
    
    if (SOC_dif2 < 0.15)          % state of charge of one actor is really low, actor is only used for steering %during driving mode
        Aeq(6,2) = 1;
    end
    
    if (SOC_dif3 < 0.15)         % state of charge of one actor is really low, actor is only used for steering %during driving mode
        Aeq(6,3) = 1;
    end
    
    if (SOC_dif4 < 0.15)          % state of charge of one actor is really low, actor is only used for steering %during driving mode
        Aeq(6,4) = 1;
    end
end
 beq = [d_1-SW_1, d_2-SW_2, d_3-SW_3, d_4-SW_4, 0, 0];
 if flag > 0                           %if it is possible the regenerative braking torque is the limit 

    
         if (Fx_v-sum(Fx(i),(1:4)) == 0) && (Fy_v-sum(Fy(i),(1:4)) == 0) && (Mz_v-sum(Mz(i),(1:4)) == 0)                                                                             %if it is not possible to follow the target trajectory the maximum of the manual braking is used
                
                
                % max. regenerative braking torque
               lb = [-M_reg_1/r_dyn_1, -M_reg_2/r_dyn_2, -M_reg_3/r_dyn_3, -M_reg_4/r_dyn_4, d_min_vl-SW_1, d_min_vr-SW_2, d_min_hl-SW_3, d_min_hr-SW_4];     %d and betato the left side are positive, to the right side are negative  %regenerative braking torque must be positive
                
                % mechanical braking
         else
            lb = [FxB_max, FxB_max, FxB_max, FxB_max, d_min_vl-SW_1, d_min_vr-SW_2, d_min_hl-SW_3, d_min_hr-SW_4];     %d and betato the left side are positive, to the right side are negative                                       %FxB_max is negative
         end                                                                         % close conditions for braking loop, max braking torque is positive

 %driving mode
else
    lb = [P_min_1, P_min_2, P_min_3, P_min_4, d_min_vl-SW_1, d_min_vr-SW_2, d_min_hl-SW_3, d_min_hr-SW_4];
end

ub = [P_max_1/(r_dyn_1*w_1), P_max_2/(r_dyn_2*w_2), P_max_3/(r_dyn_3*w_3), P_max_4/(r_dyn_4*w_4), d_max_vl-SW_1, d_max_vr-SW_2, d_max_hl-SW_3, d_max_hr-SW_4];



nonlcon = @circlecon;
x = fmincon(fun,x0,A,b,Aeq,beq,lb,ub,nonlcon);

with this following nonlinear constraints

function [c,ceq] = circlecon(x)
coder.extrinsic('ParameterSarah')
ParameterSarah;
ceq =[Fy_v-sin(d_1)*x(1)-cos(d_1)*x(5)*c_w+sin(d_2)*x(2)-cos(d_2)*x(6)*c_w+sin(d_3)*x(3)-cos(d_3)*x(7)*c_w+sin(d_4)*x(4)-cos(d_4)*x(8)*c_w; Mz_v-(-w_vl)*(cos(d_1)*x(1)-sin(d_1)*x(5)*c_w)+w_vr*(cos(d_2)*x(2)-sin(d_2)*x(6)*c_w)-w_hl*(cos(d_3)*x(3)-sin(d_3)*x(7)*c_w)+w_hr*(cos(d_4)*x(4)-sin(d_4)*x(8)*c_w)+l_v*(sin(d_1)*x(1)-cos(d_1)*x(5)*c_w+sin(d_2)*x(2)-cos(d_2)*x(6)*c_w)-l_h*(sin(d_3)*x(3)-cos(d_3)*x(7)*c_w+sin(d_4)*x(4)*c_w-cos(d_4)*x(8)*c_w); Fx_v-(x(1)/cos(d_1))-c_w*sin(d_1)*x(5)+(x(2)/cos(d_2))-c_w*sin(d_2)*x(6)+(x(3)/cos(d_3))-c_w*sin(d_3)*x(7)+(x(4)/cos(d_4))-c_w*sin(d_4)*x(8)];

c = [];

end

but I get the error

Converged to an infeasible point.Converged to an infeasible point.

fmincon stopped because the size of the current step is less than

the value of the step size tolerance but constraints are not

satisfied to within the value of the constraint tolerance.

<stopping criteria details>

Can someone help me whats the problem?

Answer 1

[Matt J]

The initial guess you have chosen

x0 = [1,1,1,1,1,1,1,1];

seems very arbitrary. You should devise an initial guess more likely to be close to expected solution values.

Also, your "nonlinear constraints" look very linear to me. Unless I'm mistaken, all the x(i) appear to be coefficients of sine and cosine terms. You should therefore use the Aeq,beq matrices, not the nonlcon input, to express them.

Finally, the abs(x(i)) expressions in your objective function are non-differentiable and will cause problems if any x(i) near zero are searched. Just impose lower bounds x(i)>=0.

[Sarah Kern]

thank you very much!