ARIMA models : plot_diagnostics, what's meaning of residuals of our model

I am studying the ARIMA models with the following tutorial: https://www.digitalocean.com/community/tutorials/a-guide-to-time-series-forecasting-with-arima-in-python-3#step-5-—-fitting-an-arima-time-series-model

After I fit the model with Step 5 — Fitting an ARIMA Time Series Model with following code:

mod = sm.tsa.statespace.SARIMAX(y,
                                order=(1, 1, 1),
                                seasonal_order=(1, 1, 1, 12),
                                enforce_stationarity=False,
                                enforce_invertibility=False)

results = mod.fit()

print(results.summary().tables[1])

and plot

results.plot_diagnostics(figsize=(15, 12))
plt.show()

I don't know the meaning: the residuals of our model are uncorrelated and normally distributed with zero-mean. I want to know what's the residual in the model, is the meaning that the residual is the difference between true value and predict value.

Why the author set the enforce_stationarity is False since the ARIMA mode need data stationarity, what's meaning of enforce_stationarity and enforce_invertibility?

 enforce_stationarity=False,
 enforce_invertibility=False

If possible, could you explain in detail. thanks!