Is the calculation of the control energy correct? Is the control energy response realistic

Dear colleagues, I am running a project to calculate the control energy dissipated by a controller for 3 inputs 3 outputs multivariable system. The multivariable controller used is Direct Nyquist Array. I managed to design a matrix of 9 decoupling compensators with different orders, some of these compensators have poles or zeros in the right hand side of s plane, but when it is multiplied with plant transfer function matrix in a close loop control, the output responses are stable, well behaved with zero steady state error. The equation used to calculate the control energy is as attached file. When I plot the output response of the Energy E, it was like unstable response, the response is shown in the attachment. Is the response correct? can such response be considered for the study saying the energy dissipation is increasing exponentially over the time? can I accept the energy response?