• Version 0.5: in development phase
  1. A first version of SQPlab for solving optimal control problems with only equality constraints has been implemented.
  2. Changes in the options
     • algo_method has been deleted and more or less replaced by algo_hessian,
     • algo_hessian specifies whether the Hessian of the Lagrangian has to be computed by the simulator (value 'Newton') or whether a quasi-Newton approximation must be computed by SQPlab (value 'quasi-Newton').
  3. Changes in the directory examples (previously called example)
     • hanging_chain is now named hanging_chain_02,
     • two more hanging chain examples have been added with the goal to illustrate how to write the simulator for an optimal control problem with only equality constraints: in hanging_chain_01 the problem is solved by the standard SQP algorithm and in hanging_chain_01b the same problem is viewed as an optimal control problem and is solved by the reduced SQP algorithm.
  4. Cosmetic improvements.
  5. All these modifications lack of intensive testing (only trials on the three hanging chain problems have been realized).
  
  
• Version 0.4.5: December 2011
  1. The hanging chain example in the example directory has been corrected.
  
  
• Version 0.4.4: February 2009
  1. Bug corrected: the unconstrained quasi-Newton version with Wolfe linesearch was not terminated ... (Thanks to Xiaohan Ma, for having mentioned that).
  2. Cosmetic improvements.
  
  
• Version 0.4.3: December 2008
  1. The Hanging Chain example has been added.
  
  
• Version 0.4.2: December 2008
  1. Bugs corrected.
  2. Cosmetic improvements.
  
  
• Version 0.4.1: October 2007
  1. Cosmetic changes.
  
  
• Version 0.4: June 2007
  1. Unconstrained problems can now be solved by a quasi-Newton algorithm using Wolfe linesearch and the inverse BFGS update formula.
  2. Important restructuration of the software.
  
  
• Version 0.3: March 2007
  1. State constraints are accepted, provided there is no inequality constraints and the Newton version of the algorithm is used.
  
  
• Version 0.2: February 2007
  1. Globalization by linesearch, using a merit function.
  2. Quasi-Newton (BFGS) approximation of the Hessian of the Lagrangian.
  
  
• Version 0.1: December 2006
  1. The local algorithms (equality and inequality constraints) using second derivatives.