The IEEE 802.15.4 Time-Slotted Channel Hopping (TSCH) is a widely used standard technology for industrial Wireless Sensor Networks (WSNs). The applications of such networks have diverse Quality-of-Service (QoS) demands that should be satisfied. Optimal configuration of the network parameters based on their QoS requirements and run-time adaptation to continuously meet the QoS requirements given all network dynamics is a great challenge for large-scale networks. The configuration space is very large for large-scale networks resulting in space exploration to be complex and extremely time-consuming. Moreover, such space exploration needs to be performed multiple times at design-time to give insight into the worst and best case performance of the mechanisms and network topology. Yet, it is needed for the run-time reconfiguration upon changes in the network. To address the stated challenges, we propose a fast and accurate enough algorithm based on Pareto algebra to extract a subset of Pareto configurations for large TSCH networks in a very short time. A proper configuration can be then picked from this set. Having such a fast optimization algorithm, the network can react to the changes in the link quality, routing topology, and reconfigure itself optimally at an appropriate time. The performance of the proposed technique is extensively evaluated, and compared with other approaches such as basic incremental Pareto analysis, and genetic algorithm, showing its superior execution time and quality of configurations in terms of accuracy and diversity.