I don't know if this will work but, watching Andy's video triggered an idea - If I use a spring loaded, normally on, SPDT switch , to allow me (in its normal position) to drive the train over the magnet, and press the switch to momentarily break the circuit and immediately spring back, the momentary break in power may simulate the "jiggle", and cause the couplers to relax and operate - I think it is certainly worth a try and have found a supplier in Cullercoats, Tyne and Wear - who is popping some in the post to me - fingers crossed.