Maybe your problem is solved now, but for the record, you say R2 is not used.
If so, the base of TR1 is floating.
Noise immunity is always an obscure problem and the first rule is that the base emitter circuit needs to be maintained at a low impedance level. Also, the triggering level of your timer needs to be set to require a certain voltage level to trigger it.
This is why logic circuits have guaranteed levels of 'maximum voltage for a 'low'; and minimum voltage for a 'high'.
Post 6 is a good way of setting the impedance to a lower value than you currently have. As well, you can use a shunt capacitance to lower the AC impedance of the base circuit.
The way TR1 is set up, the input voltage required to trigger is only about 0.5 volt. Also, I believe the odd half volt pulse is generated from capacitive coupling between the lock and the unlock wires in the cable harness.
To set the triggering level for the timer at say 6 volt, then use a base to emitter shunt resistor of 4K7 ohm. Use say 1 Kohm if you want to raise the triggering point even higher.
From what you present, R2 is redundant anyway. Maybe even wire up TR1 like you have the base of TR3 wired.