The prediction accuracies resulted from random forest regression models change each time I run the model

Every time I run the RF model from the begining I got different accuracies I have run the following code:

df17_tmp1 = df17_tmp.sample(frac=6, replace = True).reset_index(drop=True)
    
x_3d = df17_tmp1[col_in_3d] # Features;  
y_3d = df17_tmp1['over/under_exc_vol(m3)'].values  # Target
   
# In[29]:
   
x_train_3d, x_test_3d, y_train_3d, y_test_3d = train_test_split(x_3d, y_3d, test_size = 0.3, random_state = 42)
   
# # train RF

# In[30]:

x_train_3d = x_train_3d.fillna(0).reset_index(drop = True)
x_test_3d = x_test_3d.fillna(0).reset_index(drop = True)

y_train_3d[np.isnan(y_train_3d)] = 0
y_test_3d[np.isnan(y_test_3d)] = 0

rf_3d = RandomForestRegressor(n_estimators = 70, random_state = 42)
rf_3d.fit(x_train_3d, y_train_3d)

# # Predict with RF and evaluate

# In[31]:

prediction_3d = rf_3d.predict(x_test_3d)
mse_3d = mean_squared_error(y_test_3d, prediction_3d)
rmse_3d = mse_3d**.5
abs_diff_3d = np.array(np.abs((y_test_3d - prediction_3d)/y_test_3d))
abs_diff_3d = abs_diff_3d[~np.isinf(abs_diff_3d)]

mape_3d = np.nanmean(abs_diff_3d)*100
accuracy_3d = 100 - mape_3d   

I got the following results in terms accuracies:

85.94 /
85.71/
85.83 /
82.64 /
86.56 /
85.24 /
83.40 /
82.39 /
84.98 /
83.81 /

So, is that normal? and which accuracy should be considered?

Although you set a random_state in your train_test_split() to generate a deterministic split and in the RandomForestRegressor()
which would control the randomness within the algorithm, the difference is occurring due to the random sampling you are applying to your dataframe here:

df17_tmp1 = df17_tmp.sample(frac=6, replace = True).reset_index(drop=True)

You should replace the above line with the following:

df17_tmp1 = df17_tmp.sample(frac=6, replace = True, random_state = 42).reset_index(drop=True)

to get the same output on every run.

Please refer to the documentation and this thread to learn more.