[BugFix]Support redundant experts in EPLB (#3473)

This PR adds support for redundant experts in the EPLB. 

Key points: 
- Use global_num_experts = num_experts + num_redundant_experts
consistently.
- Backward compatible when num_redundant_experts=0. 

Tested 
On a 16-rank setup (W8A8) with static EPLB and expert_map_path,
verifying router logits shape and successful requests.

- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0

Signed-off-by: yechao237 <yechao20180411@gmail.com>

vllm_ascend/torchair/ops/torchair_fused_moe.py

@@ -856,8 +856,9 @@ class TorchairAscendUnquantizedFusedMoEMethod(UnquantizedFusedMoEMethod):
         shared_experts: Optional[Any] = None,
         **kwargs,
     ) -> torch.Tensor:
-
-        is_deepseek_v3_r1 = global_num_experts == 256
+        global_redundant_expert_num = get_ascend_config(
+        ).init_redundancy_expert
+        is_deepseek_v3_r1 = global_num_experts - global_redundant_expert_num == 256
         # NOTE: now npu_moe_gating_top_k can only support `group_count=256` pattern
         if is_deepseek_v3_r1:
             topk_weights, topk_ids, _ = torch_npu.npu_moe_gating_top_k(