diff --git a/mindspore/nn/layer/normalization.py b/mindspore/nn/layer/normalization.py
index 65df2d78ed0..fc08744ccdc 100644
--- a/mindspore/nn/layer/normalization.py
+++ b/mindspore/nn/layer/normalization.py
@@ -149,6 +149,8 @@ class _BatchNorm(Cell):
     def construct(self, x):
         if self.input_dims == '2d':
             _shape_check(self.shape(x))
+        if self.input_dims == '1d':
+            _shape_check_2d(self.shape(x))
         if self.use_batch_statistics is None:
             flag = self.training
         else:
@@ -200,6 +202,12 @@ def _channel_check(channel, num_channel):
         raise ValueError("the input channel is not equal with num_channel")
 
 
+@constexpr
+def _shape_check_2d(input_shape):
+    if len(input_shape) != 2:
+        raise ValueError("The input must has 2 dims.")
+
+
 @constexpr
 def _shape_check(in_shape):
     if len(in_shape) != 4:
diff --git a/mindspore/ops/_grad/grad_math_ops.py b/mindspore/ops/_grad/grad_math_ops.py
index ae0b06745e3..bae9034c620 100755
--- a/mindspore/ops/_grad/grad_math_ops.py
+++ b/mindspore/ops/_grad/grad_math_ops.py
@@ -980,7 +980,8 @@ def get_bprop_scalar_accumulatenv2(self):
         dx = ()
         for _ in range(len(x)):
             dx = dx + (dout,)
-        return dx
+        return (dx,)
+
     return bprop
 
 
@@ -992,7 +993,7 @@ def get_bprop_scalar_addn(self):
         dx = ()
         for _ in range(len(x)):
             dx = dx + (dout,)
-        return dx
+        return (dx,)
 
     return bprop
 
diff --git a/tests/ut/python/ops/test_ops.py b/tests/ut/python/ops/test_ops.py
index ef5b0953362..debc47185eb 100755
--- a/tests/ut/python/ops/test_ops.py
+++ b/tests/ut/python/ops/test_ops.py
@@ -1671,13 +1671,11 @@ test_case_array_ops = [
     ('AddN', {
         'block': NetForTupleInput(P.AddN()),
         'desc_inputs': [[2, 3, 3, 5], [2, 3, 3, 5]],
-        'desc_bprop': [[2, 3, 3, 5]],
-        'skip': ['backward']}),
+        'desc_bprop': [[2, 3, 3, 5]]}),
     ('AccumulateNV2', {
         'block': NetForTupleInput(P.AccumulateNV2()),
         'desc_inputs': [[2, 3, 3, 5], [2, 3, 3, 5]],
-        'desc_bprop': [[2, 3, 3, 5]],
-        'skip': ['backward']}),
+        'desc_bprop': [[2, 3, 3, 5]]}),
     ('Shape', {
         'block': P.Shape(),
         'desc_inputs': [[3, 3, 2, 2]],
diff --git a/tests/ut/python/pynative_mode/nn/test_batchnorm.py b/tests/ut/python/pynative_mode/nn/test_batchnorm.py
index 08f84f2fe36..a0fda8bd5e2 100644
--- a/tests/ut/python/pynative_mode/nn/test_batchnorm.py
+++ b/tests/ut/python/pynative_mode/nn/test_batchnorm.py
@@ -67,10 +67,10 @@ def test_bn2d():
 def test_bn1d():
     """ut of nn.BatchNorm1d"""
     bn = nn.BatchNorm1d(3)
-    input_data = Tensor(np.random.randint(0, 1, [1, 3, 100, 100]).astype(np.float32))
+    input_data = Tensor(np.random.randint(0, 1, [1, 3]).astype(np.float32))
     output = bn(input_data)
     output_np = output.asnumpy()
-    assert isinstance(output_np[0][0][0][0], (np.float32, np.float64))
+    assert isinstance(output_np[0][0], (np.float32, np.float64))
 
 
 def test_bn2d_train():