Neural Network-Based Beam Pumper Model Optimization

Table 2

Predicted optimization parameters.

Schemes	Polished rod load	Peak torque of reducer	Maximum negative torque	Root-mean-square matrix
Preoptimization	6.325 × 10^4 N	4.015 × 10^4 Nm	1.367 × 10^4 Nm	1.846 × 10^4 Nm
Scheme 1	6.271 × 10^4 N	3.625 × 10^4 Nm	1.146 × 10^4 Nm	1.638 × 10^4 Nm
Scheme 2	6.428 × 10^4 N	2.738 × 10^4 Nm	1.085 × 10^4 Nm	1.148 × 10^4 Nm
Scheme 3	5.649 × 10^4 N	2.817 × 10^4 Nm	0.64 × 10^4 Nm	1.075 × 10^4 Nm
Schemes	Cyclic load coefficient	Rated power of motor	Stroke	Stroke number
Preoptimization	1.625	13.25 kW	4 m	5
Scheme 1	1.473	12.47 kW	2 m	3
Scheme 2	1.395	8.47 kW	2 m	3
Scheme 3	1.42	7.49 kW	4 m	5