The escalating integration of distributed renewable energy into the grid has precipitated a decentralized and network-wide proliferation of harmonic sources. In urban distribution networks, the pronounced capacitive and distributed-parameter characteristics of cable lines exacerbate harmonic complexity, rendering them predisposed to multipoint local resonance. This study proposes a rigorous harmonic resonance suppression strategy for cabled distribution networks, grounded in global and local collaborative optimization. The global controller employs modal analysis to discern network-wide resonance modes, formulating a global admittance optimization reshaping model for weighted harmonic voltage to quell resonance across the entire network. Addressing dynamic changes in network parameters, a network division predicated on an event-triggered mechanism determines the local control scope for each suppression device. Ultimately, a localized dual-objective dynamic optimization adjustment model reshapes the virtual conductance of voltage-regulating active power filters and multifunctional grid-connected inverters in each region. Results attest to the efficacious suppression of harmonic resonance voltages to a qualified level, coupled with the successful resolution of parameter disturbances through local conductance adjustment.

As new energies integrated into grid, the harmonics in the grid are more serious, and the Voltage Detection Active Power Filter (VDAPF) has been used as the mitigation equipment to solve the harmonic problems in power electronic distribution network. The global optimization method generally utilizes 15 min to mitigate harmonic which is named long-time-scale, and it is impossible to mitigate harmonic timely. This paper studies a multi-time-scale harmonic mitigation method which adds 5 min prediction to the global optimization method which is named multi-time-scale. The objective is to correct the deviation of harmonic caused by the long-time-scale harmonic prediction error. This paper utilizes MPC to predict the amplitude and phase variation of the harmonics in short time scale, and utilizes distributed VDAPF to optimize harmonic in long time scale. The IEEE 33 nodes system was used as an example to conduct a comparative analysis of mitigation effects in two different harmonic injection scenarios which injected separately 10 and 20%. The results shows that the mitigating effect of the global optimization method is limited under two different harmonic injection scenarios, and it needs to combine with short time scale harmonic prediction information to ensure the effectiveness and rationality of the mitigation results. When the harmonic current fluctuates, there are unqualified voltage distortion nodes in each region, and the voltage distortion rate is larger. But the MPC can change harmonic prediction in a short time, so every harmonic of each VDAPF conductance values is set bigger, and harmonic conductance of the higher harmonics is relatively lower.