GA optimisation, not finding the fittest and stalls

Im trying to use the GA optimisation tool to find the best place to put sensors on plate. all feasible locations are loaded as a file and the GA would vary the indices that correspond to x,y coordinates. I have a function that then calculates the fitness for the give sensor position.

I start with a random initial population, with 500 generation 150 population size and 70 stall limit: after a few generations it does find a 'fitter' child. But I am able to improve this manually by changing the indecisive and evaluating fitness. Also starting with 149 random and one that is 'fitter' but not the 'fittest', the GA just stalls on the 'fitter' and doesn't improve it.

It always stalls and im not clear what the problem is as the terminating statement is something along the lines of, change less than tol fun or stall generation exceed.

Have tried reducing the tolfun, increasing the stall limit and also just multiplying the fitness function by 100. What is the difference between the two tolfun and stall generation limit, they both have the exact same definition in the manual.

After 200 stalled generations it gives the following message Optimization terminated: average change in the penalty fitness value less than options.TolFun and constraint violation is less than options.TolCon.